BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//wp-events-plugin.com//7.2.3.1//EN
TZID:Asia/Kolkata
X-WR-TIMEZONE:Asia/Kolkata
BEGIN:VEVENT
UID:168@cds.iisc.ac.in
DTSTART;TZID=Asia/Kolkata:20251222T160000
DTEND;TZID=Asia/Kolkata:20251222T170000
DTSTAMP:20251219T071658Z
URL:https://cds.iisc.ac.in/events/m-tech-research-thesis-colloquium-cds-an
 -explainable-adaptive-learning-framework-for-brain-tissue-extraction-from-
 mri/
SUMMARY:M.Tech Research Thesis {Colloquium}: CDS: "An Explainable Adaptive 
 Learning Framework for Brain Tissue Extraction from MRI"
DESCRIPTION:\n\nDEPARTMENT OF COMPUTATIONAL AND DATA SCIENCES\nM.Tech Resea
 rch Thesis Colloquium\n\n\n\nSpeaker : Mr. Akhil\nS.R. Number : 06-18-01-1
 0-22-23-1-23206\nTitle : "An Explainable Adaptive Learning Framework for B
 rain Tissue Extraction from MRI"\nResearch Supervisor : Prof. Debnath Pal\
 nDate &amp\; Time : December 22\, 2025\, 04.00 PM\nVenue: The Thesis Coll
 oquium will be held on HYBRID Mode\n# 102 CDS Seminar Hall /MICROSOFT TE
 AMS\nPlease click on the following link to join the Thesis Colloquium:\nMS
  Teams link\n\n\n\nABSTRACT\n\nWhile Deep Learning currently dominates neu
 roimaging\, these models suffer from a critical\, often overlooked scalabi
 lity bottleneck: model drift caused by variations in scanner physics and 
 subject positioning\, which renders them unstable on unseen datasets. To a
 ddress this limitation\, we adopted a systematic approach in this thesis\,
  moving beyond the limitations of opaque "black box" neural networks to en
 gineer a transparent\, patient-diverse solution. Our investigation progres
 sed through a rigorous analysis of failure modes\, identifying that early 
 intensity-based methods failed on complex structures where non-brain tissu
 es share the same profile as brain tissue \, motivating a shift toward tex
 ture-based feature extraction. However\, intermediate multi-classifiers tr
 ained on central slices proved inadequate\, misclassifying valid brain tis
 sue. These key learnings allowed a refinement of the approach to an adapti
 ve learning framework centered on a "Smart Anchor" strategy that dynamical
 ly selects the optimal reference slice rather than assuming a fixed centre
 . By incorporating robust preprocessing and post-processing techniques for
  inter-slice consistency and a Two-Specialist strategy that decouples core
  identification from boundary refinement\, we eliminated error propagation
  while ensuring interpretability. This systematic development yielded a ro
 bust pipeline that achieves a Dice score of 93 ± 2%\, demonstrating a hig
 h-accuracy\, clinically scalable brain tissue extraction technique through
  reproducible design that allows explainability.\n\n\n\nALL ARE WELCOME\n\
 n
CATEGORIES:Events,MTech Research Thesis Colloquium
END:VEVENT
BEGIN:VTIMEZONE
TZID:Asia/Kolkata
X-LIC-LOCATION:Asia/Kolkata
BEGIN:STANDARD
DTSTART:20241222T160000
TZOFFSETFROM:+0530
TZOFFSETTO:+0530
TZNAME:IST
END:STANDARD
END:VTIMEZONE
END:VCALENDAR