A Survey of Time Series Classification on Breast Cancer Early Detection

PhD Qualifying Examination


Title: "A Survey of Time Series Classification on Breast Cancer Early 
Detection"

by

Mr. Haoren ZHU


Abstract:

Breast cancer has become the second leading cause of women’s cancer death 
in the worldwide and has aroused growing attention in the society. Despite 
the consensus that breast cancer early detection can significantly reduce 
the treatment difficulty and cancer mortality, few people are aware of its 
necessity since most high-accuracy detection techniques are expensive and 
inconvenient to the patients. To mitigate these challenges, a solution 
recently proposed is to utilize pairs of wearable sensors to measure the 
thermal environment of the breast surface, based on which time series 
classification (TSC) can be applied to diagnose breast cancer in its early 
stage.

In this manuscript, we comprehensively investigate how TSC methods can be 
incorporated to tackle the breast cancer early detection problem. Firstly, 
we systematically review the existing TSC methods on similar time series 
data. Considering the special needs of medical scenarios and the 
functionalities of the reviewed methods, we summarize the main challenges 
of TSC on breast thermal time series data as follows: (1) noisy sensory 
data, (2) small supervised dataset, and (3) non-explainable design and 
outcome. Finally, we identify and discuss the potential directions to 
tailor TSC methods to breast cancer early detection.


Date:  			Thursday, 12 May 2022

Time:                  	4:00pm - 6:00pm

Zoom Meeting: 
https://hkust.zoom.us/j/96528616751?pwd=WEczMzZJa0RYb0c5VTVULzB4dTlnUT09

Committee Members:	Prof. Dik-Lun Lee (Supervisor)
 			Dr. Wilfred Ng (Chairperson)
 			Prof. Ke Yi
 			Prof. Xiaofang Zhou


**** ALL are Welcome ****