Rank Induction for Multiple Instance Learning

Official repository for: Kim et al., "Ranking-Aware Multiple Instance Learning for Histopathology Slide Classification: Development and Validation Study"

🧠 Overview

Rank Induction is a training strategy for Multiple Instance Learning (MIL) that leverages expert annotations using a ranking constraint—rather than exact attention matching—to guide the model’s focus toward diagnostically meaningful areas. By ranking annotated lesion patches higher than non-lesion patches, our method guides the model to focus on diagnostically meaningful regions without over-constraining attention distribution.

⚠️ Most MIL methods either ignore expert annotations or enforce overly strict attention constraints.
✅ Our method strikes a balance by ranking annotated lesion patches higher than non-lesion ones—offering better interpretability and performance, especially in low-data regimes.

Figure 1. Model performance under data scarcity

Figure 2. Model performance interpretation.

---

🔬 Key idea

Most classic MIL methods only assume that the presence of at least one positive patch makes a slide positive, with no guidance on how important each patch is. In contrast, Rank Induction uses expert annotations (patches that are known to be lesion or non-lesion) to induce a pairwise ranking preference:

• Lesion patches should have higher attention scores than
• Non-lesion patches.

Formally, for lesion patches ( s_i ) and non-lesion patches ( s_j ), we want:

$$s_i > s_j \quad \text{for all} \; (i, j) \;\text{where} \; y_i = 1, \; y_j = 0.$$

Pairwise Probability

To implement this, we convert the score difference into a pairwise probability ( P_{i,j} ):

$$P_{i,j} \;=\; \frac{1}{1 + \exp\big[-\sigma ( s_i - s_j - m )\big]},$$

where

• $s_i$ and $s_j$ are the raw attention score (before softmax),
• $\sigma$ is a scaling factor,
• $m$ is a margin to encourage a significant gap between lesion and non-lesion attention scores.

Rank Loss

We define the Rank Loss $\mathcal{L}\mathrm{rank}$ by comparing the predicted pairwise probability ($P{i,j}$) against the ground-truth preference ($\bar{P}_{i,j} \in {0,1} $) (which indicates which patch should be ranked higher):

$$\mathcal{L}_{\text{rank}} = \frac{1}{|\mathcal{P}|} \sum_{(i,j)\in \mathcal{P}} -\bar{P}_{i,j} \,\log P_{i,j} \;-\; \bigl(1 - \bar{P}_{i,j}\bigr)\,\log\bigl(1 - P_{i,j}\bigr),$$

where ($\mathcal{P}$) contains all valid (lesion–non-lesion) patch index pairs.
Finally, the slide-level classification objective (e.g., binary cross-entropy) and the rank loss are jointly optimized.

---

📊 Results

Camelyon16

Method	AUROC	AUPRC
AB-MIL	0.741 ± 0.146	0.730 ± 0.183
Attention Induction	0.742 ± 0.142	0.727 ± 0.179
CLAM-SB	0.732 ± 0.137	0.700 ± 0.179
CLAM-MB	0.794 ± 0.128	0.767 ± 0.168
DS-MIL	0.690 ± 0.170	0.666 ± 0.223
HIPT	0.483 ± 0.103	0.416 ± 0.136
Rank Induction	0.839 ± 0.050	0.850 ± 0.042

DigestPath

Method	AUROC	AUPRC
AB-MIL	0.993 ± 0.003	0.990 ± 0.005
Attention Induction	0.994 ± 0.002	0.990 ± 0.003
CLAM-SB	0.977 ± 0.019	0.965 ± 0.026
CLAM-MB	0.976 ± 0.019	0.963 ± 0.026
DS-MIL	0.995 ± 0.004	0.993 ± 0.006
HIPT	0.962 ± 0.058	0.947 ± 0.091
Rank Induction	0.995 ± 0.002	0.993 ± 0.002

SMF-stomach

Method	AUROC	AUPRC
AB-MIL	0.864 ± 0.031	0.928 ± 0.016
Attention Induction	0.868 ± 0.031	0.930 ± 0.016
CLAM-SB	0.837 ± 0.023	0.919 ± 0.013
CLAM-MB	0.838 ± 0.032	0.916 ± 0.025
DS-MIL	0.839 ± 0.031	0.915 ± 0.016
HIPT	0.676 ± 0.064	0.751 ± 0.060
Rank Induction	0.875 ± 0.008	0.937 ± 0.006

👨‍🔬 Authors

Name	ORCID	Email	Affiliation	Notes
Ho Heon Kim	0000-0001-7260-7504	hoheon0509@mf.seegene.com	$^{1}$ AI Research Center, Seegene Medical Foundation	Contributed equally
Gisu Hwang	0000-0003-1046-9286	gshwang@mf.seegene.com	$^{1}$ AI Research Center, Seegene Medical Foundation	Contributed equally
Won Chang Jeong	0009-0008-1931-5957	jeongwonchan53@mf.seegene.com	$^{1}$ AI Research Center, Seegene Medical Foundation
YoungSin Ko	0000-0003-1319-4847	noteasy@mf.seegene.com	$^{1,2}$ AI Research Center / Pathology Center, Seegene Medical Foundation	Corresponding author

📍 Affiliations

• $^{1}$ AI Research Center, Seegene Medical Foundation, 288 Dapsimni-ro, Seoul, South Korea
• $^{2}$ Pathology Center, Seegene Medical Foundation, 288 Dapsimni-ro, Seoul, South Korea