Pramit Ghosh

Blood smear analyzer for white blood cell counting

Graphical abstractDisplay Omitted Total count and differential count of leukocytes or white blood cells (WBC) in blood samples are very important pathological factors for diagnosing a disease. There are not enough pathological infrastructures in the remote places of India and other developing countries. The objective of this work is to design a system, compatible with telemedicine, for automatic calculation of the total count and differential count of WBC from the blood smear slides. Hemocytometer based WBC counting provides more accurate result than manual counting, but hemocytometer preparation process needs expertise. As this device is targeted for remote places, blood smear technique is adopted to reduce the overhead of the operator. In the proposed system, microscopic images of blood smear sample are processed to highlight the WBC for segmentation. Region segmentation procedure involves background scaling and redundant region elimination from the region set. After segmentation, the more accurate region boundary is restored by using gradient based region growing with neighbourhood influence. Individual regions are separately classified on the basis of shape, size, color and texture features independently using different fuzzy and non-fuzzy techniques. A final decision is taken by combining these classification results, which is a kind of hybridization. A set of rules has been generated for making final classification decision based on outputs from various classifiers. The sensitivity and specificity of the system are found to be 96.4% and 79.6%, respectively on a database of 150 blood smear slides collected from different health centres of Kolkata Municipal Corporation, Kolkata, India.

Medical Aid for Automatic Detection of Malaria

The analysis and counting of blood cells in a microscope image can provide useful information concerning to the health of a person. In particular, morphological analysis of red blood cell’s deformations can effectively detect important disease like malaria. Blood images, obtained by the microscope, which is coupled with a digital camera, are analyzed by the computer for diagnosis or can be transmitted easily to clinical centers than liquid blood samples. Automatic analysis system for the presence of Plasmodium in microscopic image of blood can greatly help pathologists and doctors that typically inspect blood films manually. Unfortunately, the analysis made by human experts is not rapid and not yet standardized due to the operators’ capabilities and tiredness. The paper shows how effectively and accurately it is possible to identify the Plasmodium in the blood film. In particular, the paper presents how to enhance the microscopic image and filter out the unnecessary segments followed by the threshold based segmentation and recognize the presence of Plasmodium. The proposed system can be deployed in the remote area as a supporting aid for telemedicine technology and only basic training is sufficient to operate it. This system achieved more than 98% accuracy for the samples collected to test this system.

Round-the-clock urine sugar monitoring system for diabetic patients

A fuzzy logic based image processing application has been developed here which noninvasively measures the blood sugar level of a person from his /her urine by noting the colour change in its reaction with Benedicts reagent and displays the result so that apart from the patient, others also get informed. This system helps a diabetic patient to regularly monitor and control his/her blood sugar level by taking appropriate dose of medicine and /or controlling diet. The measurement accuracy of the method is 96.93%.

An automated system for segmenting platelets from microscopic images of blood cells

Platelet count is a very important diagnosis test to identify diseases like Dengue, Malaria, Yellow fever, and others. For dengue patient monitoring, platelet count is often needed. A suspected dengue patient needs a very quick diagnosis to give an accurate result of how critical is the condition of the patient. In most laboratories, Leishmans stained blood slides are used to count platelets. However, these manual platelet counting requires expert lab technician and the overhead increases manifold when huge blood samples are to be checked by lab technicians that make the entire process time-consuming. So, in this paper, an automated system is proposed which can extract platelets from the microscopic image of blood cells, and that makes platelet counting task easy. Microscopic images of stained blood slides are captured using a light microscope. Then using color-based segmentation and morphological operation, platelets can be extracted. A comparative study between the platelet counts obtained before and after segmentation along with calculation of the efficiency of the system has shown this method to be robust and effective for automation of platelet count system.

Child Obesity Control through Computer Game

Due to urban life style and nuclear family culture, overweight or obesity becomes an epidemic for the future generation, i.e. children. Due to fast lifestyle parents have not enough time to play with their children. As a consequence children are addicted with TV programs or computer games to fulfil their fantasy. The future generation of the society is not at all interested on the physical exercise, which causes the obesity, because extra calorie is deposited on the fat cells. This game addiction cannot be removed. The objective of this work is to control child overweight or obesity by means of introducing physical activity into computer or video game. The proposed system achieved 96% accuracy for the test sets used to test this system.

A smart phone based app for automated segmentation and counting of platelets

Different diseases like dengue, malaria, yellow fever etc. can be diagnosed by the counting of Platelets in human blood. Often platelet count is done by expert lab technicians but the problem arises when huge number of blood samples are to be tested by them. In this work we have developed an android based mobile app which takes the microscopic image of the blood smear as input and gives the total number of platelets present in the input image as output. Our system eliminates the need of expert lab technicians and also minimizes the errors that arise due to manual inspection. A comparative study between platelet count obtained by manual inspection and the platelet count given by our app shows that the present system is robust and effective for automated platelet counting.

A hybrid approach to diagnosis of tuberculosis from sputum

This paper describes a hybrid approach to diagnosis of Tuberculosis from microscopic images of sputum. It is non-invasive in nature. In Sputum microscopy analysis, Ziehl-Neelsen-stained sputum smears are examined using a compound microscope for Mycobacterium Tuberculosis (TB). TB bacteria appear in pinkish violet color and rest of the slide looks bluish, when seen through a compound digital microscope. This manual process is being automated in this work. In microscopic analysis of sputum, white and blue regions are suppressed consecutively to highlight the pink region using RGB color information. After that the pink regions are segmented using a threshold value obtained by iterative method. After segmentation, redundant or false contours are removed. A new approach to detect more accurate contour boundary has been proposed here. In this technique, once the initial contour is selected then a gradient based region growing technique is used locally to find more accurate contour of Mycobacterium Tuberculosis. Shape, color and granularity features of the suspected region of the microscopic images of sputum smears are computed for the diagnosis of TB. Using each of these features, a fuzzy decision about the presence of Mycobacterium Tuberculosis is generated. All these three individual decisions are combined to produce the final decision. This approach is characterized as hybrid as it uses both crisp and fuzzy data representation. This approach reduces the false positive and false negative rate and the sensitivity and Specificity of the system are 93.9% and 88.2% respectively, which are quite high.

An automated system for measuring hematocrit level of human blood from total RBC count

Hematocrit level measures the percentage of red blood cells in the blood. The measurement of hematocrit level plays an important diagnostic role in many diseases like anemia, bone marrow disorder, colon cancer, ulcer, etc. The conventional process of measuring the hematocrit level in pathological laboratories is tiresome and prone to manual errors. In this work, we have developed an automated system for measuring the hematocrit level from the total red blood cell count in blood. A comparison between the hematocrit level obtained by the traditional method and the one given by our system has also been included to estimate the appropriateness of the developed system. The hematocrit level calculated by our automated system is approximately the same as the one obtained from the laboratories in most of the cases. Thus, this system could overcome the drawbacks existing in the manual process and hence measure the hematocrit level more accurately.

An Automatic Non-invasive System for Diagnosis of Tuberculosis

Tuberculosis (TB) is one of the major infectious diseases in the third world countries especially African countries. TB is a curable disease if proper treatment starts on time. Sometimes, it becomes difficult to arrange sufficient infrastructure for diagnosis of the disease in the remote areas of third world countries. Telemedicine partially solves the lack of physician and technicians in the pathological laboratories. The objective of this work is to design a complete setup that can be deployed at the remote areas of the country as a supporting aid of a telemedicine system for the diagnosis of TB. This system performs analysis of images of sputum and cough of patients. Microscopic image of TB bacteria which are stained with pink color is the main criterion for detection of TB disease. This system is still in development stage. A sufficient amount of data are needed for validation of the system.

A Novel Approach for Segmenting WBCs in Smartphones Using Color-Based Segmentation

White Blood cell (WBC) is one of the most important blood cells in our body. This cell controls human immune system. The high count and low count of WBC reflect different kinds of disease in our body. Therefore, WBC counting is very important to diagnose several diseases. The pathological process of WBC counting is very time consuming, tedious, and prone to human error. Thus, an automated WBC counting system is highly needed to obtain accurate results. WBC segmentation is very important to make an automated WBC counting system. In our work, we are proposing a color-based WBC segmentation algorithm, which segments WBCs from microscopic images of peripheral blood smear in smart phones. Experimental results show the system to be robust and effective for WBC segmentation. The proposed technique will be very helpful to make an automated WBC counting system and could potentially overcome the errors arising due to manual inspection.