Image Analysis and Image Classifier Using Neural Network with Machine Learning to Perform Differential Leucocyte Count

Journal Title: Indian Journal of Pathology: Research and Practice - Year 2018, Vol 7, Issue 6

Abstract

Introduction: A five part differential leucocyte count is provided by a hematology analyser with abnormalities being detected as flags which are then later on reviewed and confirmed by a pathology trainee or pathologist based on the difficulty level of that individual case or knowledge of the particular trainee. Peripheral smear examination is considered the gold standard and is time consuming and subjective. Methods: Using a self prepared 1500 leishman stained leucocyte image dataset the machine learning programme tensor flow from google using image intensity, histogram and convolutional neural network was trained and this was put to test on 80 random leucocyte images. Results: Only 65% concordance was obtained on 5 tier leucocyte differential count. However 95% concordance was achieved by using a two tier leucocyte differential classification of polynuclear and mononuclear cells. Conclusion: Larger dataset of images are needed before image analysis using this model can be used routinely to substitute or as addon to routine peripheral smear examination.

Authors and Affiliations

Swaroop Raj B. V.

Keywords

Related Articles

Mucoepidermoid Carcinoma of Salivary Gland: An Elusive Cytology

Context: Mucoepidermoid carcinoma (MEC) is the most common malignancy of salivary glands accounting for 5–10% being more common in the parotid gland, followed by minor salivary glands, especially palate. It is one of the...

Histopathological Study of Surface Epithelial Tumours of Ovary at a Tertiary Care Hospital

Background: Ovarian cancer accounts for 3% of all cancers in females. Ovarian Tumours are most common tumours of female genital tract. Ovarian cancer is leading cause of deaths among gynaecological cancers. The aim of th...

Correlation between Clinical Diagnosis, Colonoscopy and Histological Findings of Colon Lesions: A Study of 60 Cases

Background: colonoscopy is the “gold standard” for diagnosis of colonic mucosal disease. Colonoscopy has greater sensitivity than barium enema or CT for colitis, polyps, and cancer Colonoscopy has the advantage of not on...

Expression of Androgen Receptor in Triple Negative Breast Cancer Biopsies and Its Correlation with Prognostic Markers

Context: Androgen receptor (AR) has been found to be associated with the occurrence and progression of breast cancer; however, their clinical significance in triple negative breast carcinoma (TNBC) cases is still a matte...

Histopathological Patterns of Gall Bladder Diseases in Routine Cholecystectomy Specimens: A Hospital-Based Study from a Rural Medical College, Uttar Pradesh

Introduction:Cholelithiasis is the most common disease affecting gallbladder. Gallstones cause several histological changes in gallbladder mucosa like acute and chronic inflammation, metaplasia, hyperplasia, calcificatio...

Download PDF file
  • EP ID EP524167
  • DOI 10.21088/ijprp.2278.148X.7618.11
  • Views 92
  • Downloads 0

How To Cite

Swaroop Raj B. V. (2018). Image Analysis and Image Classifier Using Neural Network with Machine Learning to Perform Differential Leucocyte Count. Indian Journal of Pathology: Research and Practice, 7(6), 772-775. https://europub.co.uk./articles/-A-524167