Abstract: A medical image processing apparatus comprises processing circuitry configured to: obtain medical image data comprising or representative of image data values and segmentation values for pixels or voxels of a volume; specify a sampling point within the volume; select a relevant pixel or voxel for the sampling point from a set of pixels or voxels neighboring the sampling point, wherein the selecting is based on a difference of an interpolated image data value at the sampling point and an image data value for each of the neighboring pixels or voxels and a distance between the sampling point and each of the neighboring pixels or voxels; and determine a segmentation value for the sampling point based on a segmentation value at a position of the selected relevant pixel or voxel.

Abstract: A camera apparatus includes one or more processors configured to determine a plurality of crop image data variation classifications representative of real-world variations in physical appearance of a crop plant as well as a surrounding area around the crop plant. Furthermore, select a first set of input color images from first training dataset comprising a plurality of different field-of-views (FOVs). Thereafter, execute plurality of different image level augmentation operations to obtain an augmented set of color images, identify and filter noisy images from a second training dataset based on a predefined set of image parameters. After that, train neural network model in a first stage on a third training dataset, re-determine new crop image data variation classifications and re-select new color images representative of the new crop image data variation classifications to further train the neural network model in a second stage to detect one or more crop plants.

Abstract: A blood cell parameter correction method includes: obtaining the optical signal information of the particles in the blood sample; according to the pulse width information in the optical signal information, dividing the particles in the blood sample to obtain particle distribution information; according to the preset correction rule, correcting the particle distribution information to obtain the corrected particle distribution information; wherein, the correction rule is related to the pulse width information in the optical signal information, and provided is also a blood sample analyzer executing the method, and storage medium storing the program executing the method.

Abstract: A microscopy system comprises a microscope and a computing device which is configured to control the microscope to capture a plurality of raw overview images showing the sample environment with different capture parameters. The computing device comprises a machine learning model trained with training raw overview images, which receives the raw overview images as input and generates therefrom an output which is or which defines the overview image.

Abstract: A classifier engine provides cell morphology identification and cell classification in computer-automated systems, methods and diagnostic tools. The classifier engine performs multispectral segmentation of thousands of cellular images acquired by a multispectral imaging flow cytometer. As a function of imaging mode, different ones of the images provide different segmentation masks for cells and subcellular parts. Using the segmentation masks, the classifier engine iteratively optimizes model fitting of different cellular parts. The resulting improved image data has increased accuracy of location of cell parts in an image and enables detection of complex cell morphologies in the image. The classifier engine provides automated ranking and selection of most discriminative shape based features for classifying cell types.

Abstract: Disclosed is a cell classification method, to be executed by an analyzer, for classifying cells contained in a specimen, including: preparing a first measurement sample by treating a specimen under a first preparation condition; obtaining a first signal from the prepared first measurement sample; classifying, by using the first signal, cells contained in the first measurement sample; preparing a second measurement sample by treating the specimen under a second preparation condition different from the first preparation condition; obtaining a second signal from the prepared second measurement sample; classifying, by using the second signal, cells contained in the second measurement sample; and comparing a result of the cell classification performed by using the first signal and a result of the cell classification performed by using the second signal, with each other, and outputting an analysis result including a number of cells on the basis of a result of the comparison.

Abstract: A device for supporting bone marrow reading based on image analysis includes an image acquisition unit for acquiring a bone marrow image of a subject for bone marrow reading by an imaging device; a discrimination index unit for classifying the acquired bone marrow image into a class among a plurality of classes corresponding to preset discrimination indices; and a result providing unit for providing the result read by the discrimination index unit.

Abstract: A computer system is configured to receive a dataset of image-derived features for a plurality of images, reduce the dimensionality of this dataset, identify clusters within the dimensionally-reduced dataset, and generate a visual representation of the datapoint of the dimensionally-reduced dataset as icons grouped by cluster. User input is received to apply user classification labels to the images for inclusion in a training dataset. A user interface is useable to present information to the user and receive information from the user to facilitate the application of user classification labels.

Abstract: Methods and devices for concentrating target cells using dielectrophoresis (DEP) are disclosed. The method allows relatively high throughput of sample through a microfluidic device in order to allow rapid capture of target cells even when they are present in low concentrations within the sample. The method utilizes multiple chambers through which samples will flow, the chambers arranged such that the first capture area has a larger area and faster flow rate than a second chamber, the second chamber being positioned downstream of the first capture area and being smaller with a slower flow rate to further concentrate the material captured in the first capture area.

Abstract: According to at least one aspect, a system configured to count cells in a vessel is provided. The system comprises an imaging system configured to image cells in the vessel and a controller coupled to the imaging system. The controller is configured to control the imaging system to capture a focused image of the cells and estimate a number of cells in the focused image. The controller is configured to control the imaging system to capture a focused image of the cells at least in part by controlling the imaging system to capture a plurality of images of the cells in a plurality of focal planes, determining an area of at least one cell in each of the plurality of images, and selecting one image from the plurality of images as the focused image using the area of the at least one cell in the plurality of images.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing fundus images using fundus image processing machine learning models.

Abstract: Implementations of the present disclosure provide a solution for image/video processing. In this solution, an intermediate image can be obtained from an LR image. Then, a smooth value associated with an image block in the intermediate image could be determined based on a gradient of a target pixel in the image block. Further, an HR image can be generated from the intermediate image based on the smooth value. In this way, the filtering process can be performed based on a smooth value of an image block, thereby improving the efficiency of image/video processing and lowering down the computational and memory costs.

Abstract: Apparatus and methods are described for use with a cell sample that includes a plurality of cells disposed in a monolayer. A series of images associated with a series of depth levels of the cell sample are acquired, by performing a depth scan of cell sample with a microscope. One of the depth levels is identified as being an optimum focal plane for imaging one or more entities within the cell sample using the microscope, by identifying that a drop in image contrast occurs at the identified depth level relative to the image contrast of images acquired at depth levels within a given range of distances from the identified depth level. The cell sample is imaged using the microscope, by focusing the microscope at an investigative depth level that is based on the identified depth level. Other applications are also described.

Abstract: Disclosed is a method for detecting cells having at least one anomaly in a cytological sample on the basis of at least one first digitised digitised-electron-microscopy image of the sample.

Abstract: Provided by the present invention is a picture focusing method, comprising: acquiring a video picture, and, according to an instruction command, determining a target object in the video picture; determining picture depth information of a video picture region in which the target object is located; acquiring parameter information of a camera corresponding to multiple video picture frames; using the picture depth information of the video picture region and the parameter information of the camera as the input of a preset focusing neural network model; and using focus adjustment information of the camera to perform a focus adjustment operation on the camera.

Abstract: An apparatus and method for generating a treatment plan for salutogenesis, the apparatus comprising a at least a processor and a memory communicatively connected to the processor, the memory containing instructions configuring the at least a processor to receive physiological data associated with a user and comprising a plurality of biomarkers, wherein the plurality of biomarkers comprise at least a glycocalyx degradation biomarker, determine a concentration for each at least a glycocalyx degradation biomarker of the plurality of biomarkers, classify the at least a glycocalyx degradation biomarker to a disease condition and a treatment label as a function of the concentration, and generate a treatment plan as a function of the disease condition and the treatment label.

Abstract: A dialysis access site monitoring system may generate a treatment recommendation for treating a condition of an access site based on a video of the access site. The dialysis access site monitoring system may include an apparatus having a processor and a memory coupled to the processor. The memory may include instructions that, when executed by the processor, may cause the processor to generate video information based on a video of a dialysis access site of a patient, determine change in the number of pixels (CNP) information of the video information, the CNP information associated with movement of a skin surface of the patient due to blood flow through the dialysis access site, determine frequency domain information of the CNP information, determine a maximum-to-median power (M2) value of the frequency domain information, and determine at least one access site characteristic based on the M2 value.

Abstract: Aneurysms are classified and quantitatively analyzed based on medical image data acquired from a subject. In general, one or more algorithms are implemented to automatically classify, or otherwise diagnose, and measure aneurysms and their change over time. These algorithms make use of artificial intelligence and deep learning to develop quantitative analytics that can be consolidated into diagnostic reports.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing fundus images using fundus image processing machine learning models.

Abstract: Automatic substance preparation and evaluation systems and methods are provided for preparing and evaluating a fluidic substance, such as e.g. a sample with bodily fluid, in a container and/or in a dispense tip. The systems and methods can detect volumes, evaluate integrities, and check particle concentrations in the container and/or the dispense tip.

Abstract: A method of determining a path of a tubular structure comprises: obtaining volumetric medical imaging data that represents anatomy of a subject including the tubular structure; performing a position-determining procedure that comprises: obtaining an initial position of a current point on said path of the tubular structure; selecting a sub-region based on said initial position; obtaining values of at least one parameter representative of the tubular structure; inputting to a trained model both data from said selected sub-region and said parameter values, the trained model having been trained to determine paths of tubular structures; outputting by the trained model a position for a next point on said path of the tubular structure; and updating said initial position to said next point; and repeating the position-determining procedure for updating the initial position thereby to obtain the path of the tubular structure.

Abstract: An embodiment for photographic capture setting modification is provided. The embodiment may include capturing user feedback of a photograph. The embodiment may include identifying user preferences for one or more device parameters based on image recognition and machine learning of the photograph and the captured user feedback, respectively. The embodiment may further include generating a corpus. The embodiment may also include, in response to determining a user is preparing to capture a photograph, deriving one or more viewfinder tags for a scene currently within a viewfinder of a capture device. The embodiment may further include identifying one or more device parameters based on a matching of the one or more derived viewfinder tags and the one or more contextual tags. The embodiment may also include, in response to determining current user settings of the one or more identified device parameters do not match the identified user preferences, performing an action.

Abstract: A model-based method of inspecting a specimen for presence of one or more interferent, such as Hemolysis, Icterus, and/or Lipemia (HI L) is provided. The method includes generating a pixelated image of the specimen in a first color space, determining color components (e.g., an a-value and a b-value) for pixels in the pixelated image, classifying of the pixels as being either liquid or non-liquid, defining one or more liquid regions based upon the pixels classified as liquid, and determining a presence of one or more interferent within the one or more liquid regions. The liquid classification is based upon a liquid classification model. Pixel classification may be based on a trained multiclass classifier. Interference level for the one or more interferent may be provided. Testing apparatus adapted to carry out the method are described, as are other aspects.

Abstract: Embodiments facilitate prediction of anti-vascular endothelial growth (anti-VEGF) therapy response in DME or RVO patients. A first set of embodiments discussed herein relates to training of a machine learning classifier to determine a prediction for response to anti-VEGF therapy based on a vascular network organization via Hough transform (VaNgOGH) descriptor generated based on FA images of tissue demonstrating DME or RVO. A second set of embodiments discussed herein relates to determination of a prediction of response to anti-VEGF therapy for a DME or RVO patient (e.g., non-rebounder vs. rebounder, response vs. non-response) based on a VaNgOGH descriptor generated based on FA imagery of the patient.

Abstract: Disclosed is a particle analysis method for identifying infections, comprising: step A of irradiating light on a measurement sample prepared by mixing a test blood specimen containing particles, collected from a subject, a fluorescent dye that stains a nucleic acid, and a hemolytic agent, thereby acquiring scattered light and fluorescence from each particle contained in the measurement sample, step B of specifying particles that are substantially not contained in blood of control subjects not suffering from infection and are determined to be neutrophils contained in blood of patients with infection, based on the scattered light and the fluorescence, and acquiring particle number information on the specified particles, and step C of determining whether the test blood specimen is a specimen collected from patients with bacterial infection or a specimen collected from patients with viral infection, based on the particle number information.

Abstract: A classifier engine provides cell morphology identification and cell classification in computer-automated systems, methods and diagnostic tools. The classifier engine performs multispectral segmentation of thousands of cellular images acquired by a multispectral imaging flow cytometer. As a function of imaging mode, different ones of the images provide different segmentation masks for cells and subcellular parts. Using the segmentation masks, the classifier engine iteratively optimizes model fitting of different cellular parts. The resulting improved image data has increased accuracy of location of cell parts in an image and enables detection of complex cell morphologies in the image. The classifier engine provides automated ranking and selection of most discriminative shape based features for classifying cell types.

Abstract: Systems and methods for generating virtually stained images of unstained samples are provided. According to an aspect of the invention, a method includes accessing an image training dataset including a plurality of image pairs. Each image pair includes a first image of an unstained first tissue sample, and a second image acquired when the first tissue sample is stained. The method also includes accessing a set of parameters for an artificial neural network, wherein the set of parameters includes weights associated with artificial neurons within the artificial neural network; training the artificial neural network by using the image training dataset and the set of parameters to adjust the weights; accessing a third image of a second tissue sample that is unstained; using the trained artificial neural network to generate a virtually stained image of the second tissue sample from the third image; and outputting the virtually stained image.

Abstract: An image processing method for determining and reporting a velocity of a plurality of objects in a space-time image includes: providing an imager to acquire a plurality of images of a line; receiving a plurality of images; generating a space-time image from the plurality of images of the line; applying a mask to a subset image of the space-time image to provide a masked subset area; generating a rotational look up table of intensity values of pixels of columns and rows of the masked subset area; generating a radon transform for each subset image; computing a velocity of a plurality of objects in the subset image based on the radon transform; repeating the step of applying a mask to the step of computing a velocity; and displaying the velocity of a plurality of objects overlaying the space-time image substantially in real-time as the plurality of images are acquired.

Abstract: Embodiments classify lung nodules by accessing a 3D radiological image of a region of tissue, the 3D image including a plurality of voxels and slices, a slice having a thickness; segmenting the nodule represented in the 3D image across contiguous slices, the nodule having a 3D volume and 3D interface, where the 3D interface includes an interface voxel; partitioning the 3D interface into a plurality of nested shells, a nested shell including a plurality of 2D slices, a 2D slice including a boundary pixel; extracting a set of intra-perinodular textural transition (Ipris) features from the 2D slices based on a normal of a boundary pixel of the 2D slices; providing the Ipris features to a machine learning classifier which computes a probability that the nodule is malignant, based, at least in part, on the set of Ipris features; and generating a classification of the nodule based on the probability.

Abstract: A system and method for collecting, transporting, and analyzing dried bodily fluid samples using a sample collection device incorporating extraction markers, an imaging device to take images of dried samples on the collection device, and a computing device to analyze data points from the images so as to measure various properties of the collected sample such as the volume of blood initially collected, and the portion of that volume containing plasma and erythrocytes.

Abstract: A method for analyzing digital holographic microscopy (DHM) data for hematology applications includes receiving a plurality of DHM images acquired using a digital holographic microscopy system. One or more connected components are identified in each of the plurality of DHM images and one or more training white blood cell images are generated from the one or more connected components. A classifier is trained to identify a plurality of white blood cell types using the one or more training white blood cell images. The classifier may be applied to a new white blood cell image to determine a plurality of probability values, each respective probability value corresponding to one of the plurality of white blood cell types. The new white blood cell image and the plurality of probability values may then be presented in a graphical user interface.

Abstract: A method comprises: receiving, via a processor, an image depicting a tissue; quantifying, via the processor, the image based on: segmenting, via the processor, the image into a plurality of segments; identifying, via the processor, a plurality of histological elements in the segments; forming, via the processor, a network graph comprising a plurality of nodes, wherein the histological elements correspond to the nodes; measuring, via the processor, a feature of the network graph; performing, via the processor, a transformation on the image based on the feature; determining, via the processor, a non-parametric feature of the image based on the transformation; saving, via the processor, the non-parametric feature onto a database.

Abstract: A system for the detection of blood vessels includes an image sensor coupled to generate video data including a sequence of images of the blood vessels, and a heart rate monitor to measure a heart rate of a patient and to generate heart rate data. A controller is coupled to the image sensor to receive the video data, and coupled to the heart rate monitor to receive the heart rate data. The controller includes logic that when executed by the controller causes the controller to perform operations including isolate localized motion of the blood vessels in the video data using the heart rate data. The controller also computes a blood vessel mask (that includes differences between the video data and the localized motion of the blood vessels) and combined video data (that includes the video data and the blood vessel mask).

Abstract: Calibration of optical computing devices is achieved using mapping functions that map real detector responses to simulated detector responses which are simulated using high-resolution spectra of traceable optical filters and optical computing device characteristics.

Abstract: A cytological image processing device, a method for quantifying characteristics of cytological image and a non-transitory computer readable medium thereof are provided. The cytological image processing device executes the method including the steps: loading a cytological image with a plurality of pixels; removing a background part from the cytological image according to at least one of a plurality of color attributes of each pixel of the cytological image in a color space to obtain a background-subtracted image; dividing the background-subtracted image into a nuclear part and a cytoplasmic part; detecting a plurality of edge pixels in the background-subtracted image to segment a plurality of nuclear blocks in the background-subtracted image; generating a parameter set according to the nuclear blocks; and generating an output image by visualizing the background-subtracted image.

Abstract: A control device defines, for three-dimensional blood vessel data of a blood vessel corresponding to a branched center line, a cylindrical coordinate system, thereby converting this into blood vessel data in an orthogonal coordinate system. The control device uses a coordinate value of the coordinate axis in a radial direction at a first coordinate value of the coordinate axis in a center line direction and a coordinate value of the coordinate axis in the radial direction at a second coordinate value of the coordinate axis in the center line direction, thereby interpolating a coordinate value of the coordinate axis in the radial direction between the first coordinate value and the second coordinate value. The control device converts, after interpolation, the blood vessel data into blood vessel data in the cylindrical coordinate system.

Abstract: Embodiments of the present invention seek to provide an accurate way to measure the length of a fish. Some embodiments of the present invention propose utilizing a known length of a marker within a camera view or picture to determine the length of the fish also shown in the same camera view or picture. This determination could be performed in real-time or later remotely. Further, the measurement of the length of the fish is reliable because it is not dependent on a zoom value or angle of the camera and there is no required measurement device. Further, the technique is easy to use and can be performed with a camera phone or tablet.

Abstract: A method and apparatus for identifying one or more target constituents (e.g., white blood cells) within a biological sample is provided. The method includes the steps of: a) adding at least one colorant to the sample; b) disposing the sample into a chamber defined by at least one transparent panel; c) creating at least one image of the sample quiescently residing within the chamber; d) identifying target constituents within the sample image; e) quantitatively analyzing at least some of the identified target constituents within the image relative to one or more predetermined quantitatively determinable features; and f) identifying at least one type of target constituent within the identified target constituents using the quantitatively determinable features.

Abstract: The disclosure concerns a method for measuring and reporting vascularity in a biological tissue sample. The method generally includes: within a digital image of a tissue section, (i) identifying endothelial cells, lymphatic cells, or a combination thereof; (ii) mapping one or more proximity regions, each of the proximity regions defining an area between detected vessels and a first distance outwardly therefrom; and (iii) calculating one or more of: a vessel proximity score or a hypoxia score, wherein the vessel proximity score relates a composition of objects within the proximity regions, and wherein the hypoxia score relates a composition of tissue within or outside of the proximity regions, respectively.

Abstract: This document describes optical central venous pressure measurement. To determine the central venous pressure (CVP) of a person optically, video of a right side of the person's neck is captured. By way of example, a medical professional records a video of the right side of the person's neck using a smartphone. The right side of the person's neck is captured because it is where the person's external and internal jugular veins are located and pulsatile motions that are usable to measure CVP occur in those veins. The video is then processed according to video motion amplification techniques to generate a reconstructed video of the right side of the person's neck. In the reconstructed video, the pulsatile motion of the person's venous system that occurs at the right side of their neck is visually amplified. Using the reconstructed video, measurements are made of a distance between a peak of the visually-amplified pulsatile motion and an anatomical feature of the person.

Abstract: In imaging analysis of a living body, a Region Of Interest (ROI) is set on the basis of the state of radiopharmaceutical accumulation. An example for setting an ROI includes: performing first transformation for anatomically standardizing, with the use of a positive template, a nuclear medicine image acquired by applying a radiopharmaceutical to a subject; performing second transformation for anatomically standardizing, with the use of a negative template, the nuclear medicine image; calculating a degree of similarity between a first anatomical standardization image acquired by the first transformation and the positive template; calculating a degree of similarity between a second anatomical standardization image acquired by the second transformation and the negative template; and applying, to an ROI template, inverse transformation of the first transformation or the second transformation, whichever has the higher of the calculated degrees of similarity, in order to set the ROI.

Abstract: A method and apparatus is provided for medical image processing. The processing of the medical image extracting a living valve image by analyzing a 3D image of an anatomical organ of a subject, specifying a distribution of an amount of calcium in the living valve image from CT pixel values in the living valve image, and specifying a distribution of difficulty levels of implanting an artificial valve to the anatomical organ based on a comparison between the amount of calcium and a predetermined value.

Abstract: The present invention relates to an ultrasound imaging system (10) for inspecting an object (97) in a volume (40). The ultrasound imaging system comprises an image processor (36) configured to conduct a segmentation (80) of the object (97) simultaneously out of three-dimensional ultrasound mage data (62) and contrast-enhanced three-dimensional ultrasound image data (60). In particular, this may be done by minimizing an energy tem taking into account both the normal three-dimensional ultrasound image data and the contrast-enhanced three-dimensional image data. By this, the normal three-dimensional ultrasound image data and the contrast-enhanced three-dimensional image data may even be registered during segmentation. Hence, this invention allows a more precise quantification of one organ in two different modalities as well as the registration of two images for simultaneous visualization.

Abstract: A device comprises: a nucleus-candidate-region extracting section that extracts, from a captured image obtained by image-capturing a sample piece including a target cell having a nucleus, a nucleus candidate region corresponding to the nucleus; a basic-probability-information acquiring section that acquires, for each of a plurality of determination subject regions determined on the basis of the nucleus candidate region extracted by the nucleus-candidate-region extracting section, basic probability information indicating probability that an image in the determination subject region is an image of the target cell, on the basis of a feature amount of the image of the determination subject region; and a probability-information calculating section that calculates probability information indicating probability that an image in a display subject region corresponding to the nucleus candidate region is the image of the target cell, on the basis of the basic probability information acquired for each of the plurality of

Abstract: An image processing device 20 acquires captured images obtained by imaging a sample including a target cell, performs machine learning based on a first image feature quantity, sets a plurality of object regions for detecting the target cell in the captured images, and displays the plurality of object regions in an order determined based on the first image feature quantity of each of the plurality of object regions. The image processing device 20 calculates a second image feature quantity in each of the plurality of object regions, sorts the plurality of displayed object regions in an order of the second image feature quantity similar to the second image feature quantity of a reference object region selected from among the plurality of object regions, and displays the plurality of object regions.

Abstract: A method for detecting a cancer cell, including: contacting in vitro a peripheral blood-containing sample including test cells and a labeling substance; allowing the test cells to incorporate the labeling substance to label the test cells; and detecting, as a cancer cell, a cell showing a higher degree of labeling with the labeling substance than that shown by a normal cell in the peripheral blood-containing sample.

Abstract: A method for determining a mean cell volume for a blood sample includes: illuminating the sample with incident light at a plurality of illumination wavelengths and obtaining a two-dimensional image of the sample at each of the plurality of illumination wavelengths; identifying a plurality of cells that appear in each of the images; for each one of the plurality of cells, determining an integrated optical density corresponding to each of the plurality of illumination wavelengths; for each one of the plurality of cells, determining a cell volume based on the integrated optical densities corresponding to each of the plurality of illumination wavelengths; and determining the mean cell volume for the blood sample from the cell volumes for each one of the plurality of cells.

Abstract: A method comprising: mapping an image onto multiple level sets connected by multiple branches, wherein each of said level sets corresponds to a predefined range of values for an attribute of said image; associating multiple pixels of said image to said multiple level sets in accordance with the value of each pixel; identifying a source level set associated with a source pixel of said multiple pixels; for each of said multiple pixels, determining that a distance between said source pixel and one of said multiple pixels is within a predefined threshold, wherein said distance is calculated as a function of a first distance between said source level set and said level set associated with said one of said multiple pixels; in one embodiment, the determined distance is applied to an image processing application to produce a processed image, and the processed image is rendered on a rendering medium.

Abstract: The Inventors disclosed an inexpensive method of determining fetal hemoglobin (HbF) in red blood cells of individuals with sickle cell anemia undergoing hydroxyurea treatment. The method is based on the specific elution and measurement of sickle hemoglobin (HbS) from red blood cells, wherein the amount of HbF may be determined according to the formula: HbF=Total hemoglobin?HbS. The method may also be used to determine the amount of HbF in individuals without sickle cell anemia by the specific elution of adult or normal hemoglobin (HbA) from red blood cells whereby the amount HbF may be determined according to the formula: HbF=Total hemoglobin?HbA. The method will also determine levels of HbS or HbA in individuals with sickle cell and without sickle cell respectively. The methods have the advantage of being easy to practice in underdeveloped countries.

Abstract: A method for analysing microbial growth on a solid culture medium, the method including obtaining image data of the solid culture medium and any microbial growth, generating an associated feature vector of values obtained by applying one or more filters to the image data, using a classifier to classify each pixel in a plurality of pixels in the image data based on the associated feature vector, analysing results of pixel classifications of each said pixel to derive a microbiological assessment of the solid culture medium and any microbial growth, and outputting the microbiological assessment.