Search

CompreHensive Digital ArchiVe of Cancer Imaging - Radiation Oncology (CHAVI-RO) is a multi-tier WEB-based medical image databank. It supports archiving de-identified radiological and clinical datasets in a relational database. A semantic relational database model is designed to accommodate imaging and treatment data of cancer patients. It aims to provide key datasets to investigate and model the use of radiological imaging data in response to radiation. This domain of research area addresses the modeling and analysis of complete treatment data of oncology patient. A DICOM viewer is integrated for reviewing the uploaded de-identified DICOM dataset. In a prototype system we carried out a pilot study with cancer data of four diseased sites, namely breast, head and neck, brain, and lung cancers. The representative dataset is used to estimate the data size of the patient. A role-based access control module is integrated with the image databank to restrict the user access limit. We also perform different types of load tests to analyze and quantify the performance of the CHAVI databank.

Pathogenesis of different diseases showed that some of them, especially thalassemia (T) and rheumatoid arthritis (RA) have an implicit association with oxidative stress and altered levels of reactive oxygen species (ROS). Introducing ROS level and the balance between ROS and antioxidants as essential metrics, an attempt was made to classify T and RA from normal individuals (treated as controls)(C) using synchronous fluorescence spectroscopy (SFS) and Raman line intensity of water. This non-invasive and label-free approach was backed up by a categorization algorithm that helped in the prediction of disease types from serum samples. The predictive system constituted principal component analysis (PCA) with four parameters, namely spectral intensity ratios of reduced nicotinamide adenine dinucleotide (NADH) to tryptophan (Trp) (NADH/Trp), kynurenine (Kyn) to tryptophan (Kyn/Trp), kynurenine to NADH (Kyn/NADH), and logarithmic changes in Raman line intensity of water (Rline), with the index headers containing the disease types. Rline has a positive correlation with both Kyn/Trp and Kyn/NADH and a negative correlation with NADH/Trp ratio, implying its direct or indirect association with oxidative stress. In addition to the classification of T, RA, and C a sub-classification of T into beta major and E-beta in their post and pre-splenectomized surgical stages could also be realized. Furthermore, receiver operating characteristic (ROC) analysis was deployed to ascertain that the misclassification error (ME) was negligible for the disease types. Graphical Abstract A schematic representation of the workflow converging into the categorical classification of disease classes.

We propose a de-identification system which runs in a standalone mode. The system takes care of the de-identification of radiation oncology patient’s clinical and annotated imaging data including RTSTRUCT, RTPLAN, and RTDOSE. The clinical data consists of diagnosis, stages, outcome, and treatment information of the patient. The imaging data could be the diagnostic, therapy planning, and verification images. Archival of the longitudinal radiation oncology verification images like cone beam CT scans along with the initial imaging and clinical data are preserved in the process. During the de-identification, the system keeps the reference of original data identity in encrypted form. These could be useful for the re-identification if necessary.

Tertiary oncology center clinicians are commonly faced with the problem of managing patients with a diagnosis of breast cancer made after lumpectomy in the Primary Health Care (PHC) setting. There are no studies or guidelines that address the further surgical management in this group of patients regarding sentinel lymph node biopsy (SLNB) and need for breast post-operative cavity excision. Prospective observational study was planned to evaluate the feasibility of SLNB and defining the need for definitive breast surgery in patients diagnosed with breast cancer after lumpectomy in PHC. The study was carried out from January 2015 to August 2017 in Tata Medical Center, India, approved by institutional review board (EC/TMC/36/14). Seventy patients who underwent lumpectomy with a definitive histological analysis of breast cancer were included in this study. Each patient had definitive breast surgery and SLNB using subareoral blue dye injection followed by validation axillary dissection. The identification rate (IR) for SLNB was 92% (64/70). The median number of SLNs removed was 2 (IQR 1, 3). There were 2 patients with false negative results resulting in false negative rate (FNR) of 11%. Overall, SLNB procedure has the sensitivity of 89%, NPV of 96%, and accuracy was 97%. Peri-areoral incision of initial surgery was associated with low IR (84%) and high FNR (33%). Final histopathology showed residual invasive cancer in 43% and ductal carcinoma in situ in 14% of patients. Among 21 patients where initial lumpectomy histopathology margin was free of cancer, residual malignancy was found in 57% of patients. Prior excision of lumps for breast cancer does not affect the accuracy of SLNB. Peri-areoral scar may be associated with high FNR and low IR, although further studies are needed to validate this statement. Definitive breast surgery is required for all patients, irrespective of initial lumpectomy histopathological margin status.

Quality Indicators for Sentinel Lymph Node Biopsy in Breast Cancer: Applicability and Clinical Relevance in a Non-screened Population: sentinel lymph node biopsy (SLNB) has replaced axillary lymph node dissection (ALND) as standard of care for management of early breast cancer. This study assessed our SLNB program against 11 published quality indicators (QIs). All breast cancer patients who underwent SLNB in our centre from June 2013–Dec 2015 were included. Clinical, pathological and follow-up data were extracted from the institutional REDCap data system. Analysis was done with SPSS 23. Following validation, 234 patients had SLNB, always performed along with primary surgery. Identification rate was 95.3% and > 1 SLN was identified in 72% of patients. SLNB positivity was 33%, of these, 100% underwent ALND. Overall 91% of QI eligible patients underwent SLNB. No ineligible patients (T4) underwent SLNB. For the patients who had radio colloid, injection criteria were met for 100%. Pathological evaluation and reporting criteria were met for 100% of patients. There were no axillary recurrences in a median follow-up of 2 years. 7.6% patients had SLN negative on frozen section but positive on final histology. 7.2% of patients with clinical negative nodes had pN2 disease in final histopathology report after surgery. Sixty percent of patients who had completion ALND had only positive SLN. This study supports the applicability of published QI of SLNB in a non-screened cohort of early breast cancer patients. Although QI were useful, modification based on patient characteristics and resource availability may be needed. These indicators can be used as audit tools to improve the overall accuracy of the procedure.

Oral squamous cell carcinoma (OSCC) diagnosis through computer vision approach is newly introduced technique in the modern diagnostic era. Mitotic cell count from related tissue histopathological images signifies the proliferative marker of cancer cell has been recognized as an essential phenomenon in diagnosis. This paper aims at developing an automated technique for accomplishing the task of mitotic cell count from related histopathological images. In this regard, a new machine learning based methodology incorporating random forest tree classifier learns over four entropy measures, fractal dimension, and seven Hu's moments based descriptors have been introduced. The performance validation summarizes that proposed methodology can detect mitotic cell efficiently from histopathological images of OSCC with 89% precision, 95% recall or sensitivity, 97.35% specificity, 96.92% accuracy, 96.45% AUC and 92% F-score measure.

Neoadjuvant chemotherapy (NACT) for locally advanced breast cancer (LABC), apart from increasing breast conservation rates, also provides an opportunity to assess tumour response to chemotherapy, with Pathological Complete Response (pCR) described as an independent prognostic factor and a surrogate marker for better outcome and survival. Our primary aim was to identify clinical and pathological factors associated with pCR following NACT in patients with LABC treated at our institution. Our secondary aim was to analyze the impact of pCR and associated factors on disease free survival (DFS) and overall survival (OS). A retrospective analysis of LABC patients treated with NACT between Jun 2011 and Dec 2013. Clinical and histological variables were analyzed for association with pCR (no invasive or in situ carcinoma in breast or axillary lymph nodes). Kaplan-Meier curves and Cox regression model was used for survival analysis. All values were twosided, and statistical significance was defined as p