Developing clinical high efficiency platforms for individualised treatment through integration ofadvanced radiation technology, quantitative imaging and molecular biology and machinelearning for treatment of cervix cancer.

Cervical cancer is the second most common cancer in India. Every year, around 80,000-90,000 women bear the burden of cervical cancer. In recent years, the use of advanced external radiation and brachytherapy techniques have improved the condition of the patients and their overall survival rates. This study proposal is an initiative with International collaborators to develop a robust model of efficient treatment delivery. In order to achieve this goal, we are making a move to integrate the knowledge of advanced technology, existing radiation treatment information, quantitative imaging, and available datasets from completed and ongoing clinical studies so that women diagnosed with Cervical cancer can be provided with highly precise treatment in a time-efficient manner. It is a retrospective study that will use existing database to develop automation and prediction tools.

Study Aims

1.      To develop and validate automated platforms for target delineation and planning for cervix cancer in time-efficient manner through

a.       Machine learning-based detection of abnormal cancerous tissues in multimodality medical diagnostic images.

b.      To train machine base systems for automated planning of external radiation and brachytherapy for gynaecological cancers.

2.      To use existing databases and radiation dose maps, imaging texture features and adverse events data for machine learning to develop “normal tissue complication plots “and to identify cervix cancer patient subgroups that may benefit from advanced radiation techniques (like proton treatment)

3.      To use advanced image texture analysis within ongoing institutional and collaborative clinical trials to identify “high-risk patient population†that may benefit from intensification of treatment in the future.