Chapter 340 - 230: An All-English Inquiry? Three Heroes Battle Jiang He

Chapter 340: Chapter 230: An All-English Inquiry? Three Heroes Battle Jiang He

Jiang He obediently took the paper and began to flip through it.

After reading for a moment, he suddenly asked, "Elder Fu, this paper on early intervention to block inflammatory factors... is Miller the first author?"

Fu Yunshu nodded. "Correct. It’s his most highly-rated paper this year. Not only did it receive funding from the National Institutes of Health, United States (NIH), but it has also been cited many times."

Jiang He: "Hmm..."

How should he put it?

As a true master of this era—2008—Jiang He was extremely sensitive to data.

’This paper in my hands... something about it feels... too smooth?’

The response curve after blocking the inflammatory factors, the control group’s survival rate in the animal model—every data point was perfect.

’It’s almost as if I wrote it myself.’

This feeling was a bit like Puyi’s back in the day.

Puyi was a master appraiser of treasures and could often tell a fake from the real thing with a single glance.

Others would ask him how he did it.

Puyi would simply reply, "It’s not quite the same as the ones I saw when I was young."

Jiang He felt the exact same way now.

The paper was too perfect.

’The data has most likely been polished, or even beautified by selectively removing unfavorable results.’

But Jiang He didn’t say anything.

’For now, it’s just a gut-level suspicion.’

’The burden of proof is on the one who makes the accusation.’

’I’ll just file this away for now.’

’It’s a good thing I’m going to the United States. The lab can’t sit idle while I’m away.’

’It wouldn’t be a bad idea to have everyone run a replication study of Professor Miller’s work and see if the data is as perfect as he claims.’

Jiang He closed the photocopied paper and said, "Professors, I’m finished."

Zhou Deming put down his teacup, a little surprised. "Finished? That quickly?"

"I skimmed for the core data and conclusion. I’ll study it more carefully when I get back," Jiang He answered truthfully.

Zhou Deming nodded, not pressing the matter further.

"Alright, you’ve looked at the theory, let’s get to the practical part. Xiaojiang, from now on, I won’t be speaking Chinese. You can treat me as one of those old American professors sitting in the front row at the conference in Baltimore."

A warmth spread through Jiang He’s chest.

’They even set up a mock exam.’

’The director has gone to a lot of trouble, and so have the old professors.’

He nodded and said, "Please, go ahead."

To replicate a genuine academic Q&A environment, Zhou Deming spoke very quickly.

Although he still had a bit of a Chinese accent, the professor was trying his best:

"Dr. Jiang, I have read your paper carefully. Your team claims to have achieved an AUC of 0.915 in the retrospective cohort. However, as a model based on multivariate analysis, how can you prove that your algorithm will not overfit when faced with external validation sets from different ethnic groups and geographical regions?"

A very sharp opening.

Jiang He replied instantly in fluent, standard English:

"Thank you for your question. Regarding the concern about overfitting, we took that into consideration from the very beginning of our model’s development. Our underlying architecture uses a regularized logistic regression with a penalty term (LASSO), combined with ten-fold cross-validation. More importantly, behind the 0.915 AUC that you see, we have eliminated the weights of indicators that rely solely on population-specific characteristics of the local cohort..."

Zhou Deming’s eyes lit up.

’What an impeccable answer.’

’The technical vocabulary flowed without hesitation, and his pronunciation even seemed more standard than my own.’

’This young man!’

Professor Fu Yunshu, who was at the side, also asked a question entirely in English:

"Dr. Jiang, returning to the mechanistic level, you just looked at Professor Miller’s paper. Professor Miller proposes that in the early stages of acute pancreatitis, blocking TNF-α (Tumor Necrosis Factor-α) and IL-6 is key to controlling the inflammatory cascade. However, in your predictive model, the weights of these classic indicators seem to have been deliberately lowered. Are you questioning the pathophysiological consensus of the Hopkins team?"

Jiang He replied unhurriedly:

"I am not questioning Professor Miller’s mechanistic research. From the perspective of the clinical time window, by the time we detect significantly elevated levels of TNF-α and IL-6 in a patient’s serum, the inflammatory cascade has already fully erupted. Making an early prediction at that point is already too late for clinical intervention..."

Another long string of rapid-fire words.

Fu Yunshu didn’t even understand a few of the words.

But he controlled his expression.

He observed the other two professors.

The other two professors were nodding slightly, looking as if they had understood everything.

So Fu Yunshu also nodded slightly.

In reality...

All three old professors had only a partial understanding, but each assumed the others had understood completely, so they could only nod along.

Guo Fengwan cleverly chose to cut in with a new topic:

"Dr. Jiang, your previous answer was excellent. However, early prediction is one thing, and clinical benefit is another. Even if your model can provide a warning for SAP 12 hours in advance, we currently have no specific drugs to block the progression of SAP, aside from fasting, fluid replacement, and inhibiting pancreatic enzyme secretion. Besides increasing the patient’s psychological burden, can an early warning truly and substantially change the patient’s outcome with the currently available treatment methods?"

This question directly challenged the ultimate clinical significance of the research.

Jiang He explained:

"Professor Guo, the cause of death in SAP is severe hypovolemia due to third-space capillary leakage, which in turn leads to Multiple Organ Dysfunction Syndrome (MODS)."

"If we can accurately predict 12 to 24 hours in advance which mild cases will become severe, we won’t have to wait until the patient shows signs of shock to begin resuscitation. We can initiate early goal-directed therapy (EGDT) for fluid resuscitation during the deceptive period when the patient’s vital signs appear stable, thereby maintaining visceral perfusion..."

Another torrent of rapid-fire words.

The professors were lost again.

It wasn’t that they were completely lost, just that their comprehension was far from complete.