Prognosis Prediction in COVID-19 Patients through Deep Feature Space Reasoning

Abstract

The COVID-19 pandemic has presented a unique challenge for physicians worldwide, as they grapple with limited data and uncertainty in diagnosing and predicting disease outcomes. In such dire circumstances, the need for innovative methods that can aid in making informed decisions with limited data is more critical than ever before. To allow prediction with limited COVID-19 data as a case study, we present a complete framework for progression and prognosis prediction in chest X-rays (CXR) through reasoning in a COVID-specific deep feature space. The proposed approach relies on a pre-trained deep learning model that has been fine-tuned specifically for COVID-19 CXRs to identify infection-sensitive features from chest radiographs. Using a neuronal attention-based mechanism, the proposed method determines dominant neural activations that lead to a feature subspace where neurons are more sensitive to COVID-related abnormalities. This process allows the input CXRs to be projected into a high-dimensional feature space where age and clinical attributes like comorbidities are associated with each CXR. The proposed method can accurately retrieve relevant cases from electronic health records (EHRs) using visual similarity, age group, and comorbidity similarities. These cases are then analyzed to gather evidence for reasoning, including diagnosis and treatment. By using a two-stage reasoning process based on the Dempster-Shafer theory of evidence, the proposed method can accurately predict the severity, progression, and prognosis of a COVID-19 patient when sufficient evidence is available. Experimental results on two large datasets show that the proposed method achieves 88% precision, 79% recall, and 83.7% F-score on the test sets.

Keywords: COVID-19; case retrieval; deep feature space reasoning; patient demographics; prognosis prediction.

Figure 1

Proposed Framework.

Figure 2

Sample scans before and after image enhancement.

Figure 3

(a) First 25 feature maps from the last convolutional layer, (b) First 25 selected feature maps overlayed on the input image.

Figure 4

Sample CXRs and their corresponding feature vectors.

Figure 5

Two-dimensional tSNE embedding of the 512-D feature space along with neighborhoods considered during relevant case retrieval. The red circle represents immediate neighbors whereas the blue circles indicate second stage neighbors. Each point represents a CXR embedding into the feature space along with the associated variables.

Figure 6

Effect of age difference in case retrieval on severity prediction performance.

Figure 7

Effect of various parameters in case retrieval on severity prediction performance.

Figure 8

Comparison of selected features for single scan.

Figure 9

Comparison of selected features for multiple scans.

Figure 10

Top 10 relevant cases retrieved for the query case (Patient age = 36 years).

Figure 11

Top 7 relevant cases retrieved for the query case (Patient age = 90 years).