In an Age of Images, Teaching Pathology by Hand

原创 2007年09月13日 15:50:00

BOSTON — The patient’s decline was ruthless. He became forgetful in his early 50s. By the time he saw a neurologist a few years later, he could not recall how to write in cursive, and confused pictures of a cookie and a whistle, a bed and a sandbox.

Skip to next paragraph
Photographs by Robert Spencer for The New York Times

Dr. Jeffrey T. Joseph, a pathologist, leading a first-year resident, Dr. Scott Wenson, above left, through a brain autopsy at the Beth Israel Deaconess Medical Center at Harvard. “It’s part diagnosis, part entertainment,” Dr. Joseph said of his neuropathology seminar.

A few years after that, he could no longer dress himself or speak fluently, and had developed muscle jerks that required medication to control. Diagnostically, he showed symptoms of several types of dementia, but no hallmarks that would narrow the cause to only one. He died in his early 60s.

On a Friday morning a week later, Dr. Jeffrey T. Joseph, a pathologist, was gazing at the patient’s brain, waiting for neurology residents to arrive and gaze with him. He pondered whether to begin the dissection with a traditional vertical slice down the height of the brain. “I might cut coronally, might cut horizontally,” he said, thoughtfully. “I haven’t decided yet.”

From a distance, brain autopsies seem an afterthought on life. Insurance does not cover them. They serve no lucrative purpose, so hospitals have a financial disincentive to do them. As a result, the field of neuropathology is shrinking, and its atrophy may diminish the entire field of neurology.

Neurology, after all, is a subscience of anatomy — the symptoms of brain disorders are directly related to the sites of lesions — yet many neurology residencies no longer require brain-cutting time. Without this experience, residents cannot lay eyes or hands on the organ that will be their lifeblood.

“Neuropathology as a freestanding subspecialty is going to disappear,” said Dr. Thomas Smith, professor of pathology and neurology at University of Massachusetts Medical School. “It’s going to be subsumed into general surgical pathology.” His department used to have two neuropathologists and a lab; now there is no lab and one doctor.

“If you don’t have a neuropathologist on your faculty teaching, it’s a big loss,” Dr. Smith added.

A few generations ago, neurologists in training spent months dissecting spinal cords and brains. “The main reason,” said Dr. Dan Press, assistant professor of neurology at Harvard Medical School (and a former student of Dr. Joseph’s, who still attends the seminars when he can), “was because there was no imaging then.”

Everything changed with the advent of magnetic resonance imaging. Now residents spend hours each week studying neuroradiology; M.R.I. and CT are the primary educational tools. As a result, they gain most of their anatomical knowledge from images. Specialty board examinations that used to be full of pathology pictures are now full of radiologic images.

“But pathology still allows a resolution that imaging never will,” Dr. Press said. “It forces you to see three-dimensionally. There’s no more concrete way to make it real. It’s the difference between looking at maps of a place and driving around it.”

Dr. Joseph has conducted a weekly neuropathology seminar at the Beth Israel Deaconess Medical Center at Harvard for more than 12 years. In the fall, he will leave Harvard to become a professor at the University of Calgary in Alberta. His seminar will move with him. “It’s part diagnosis, part entertainment,” Dr. Joseph said. “I like to make them look. The brain is like a geode — you don’t know what’s inside.”

A lively 52-year-old with sideburns and a Mephistophelean glint, Dr. Joseph wears a bow tie covered with dragons, or paramecia; he does not know which, because his wife — a neurologist who attended his seminar during her residency — buys them.

The autopsy suite is in the basement of the hospital, behind an unmarked door with a discreet posting: “Funeral directors, please sign the log book.” Unlike the surgical pathology suite, where there are five high-grade cutting stations, the atmosphere here is left-behind; ventilation is old and loud, and sometimes it can be hard to find a pair of sharp scissors. Bach used to play in the background, until a senior pathologist remonstrated Dr. Joseph that the autopsy suite was not a place for music; it was sacred ground, a church. Dr. Joseph pointed out that Bach was often played in churches. The argument did not prevail.

Residents drift in — neurology tuning forks and reflex hammers poking from white pockets — and move toward the dissecting table, where blades and tweezers have been laid out beside several bulging towels in pans. Dr. Press is here, too; he was one of the physicians caring for the patient. Dr. Joseph unwraps one of the towels. It is the control case, a normal brain from a patient who died of unrelated illness.

“Nice and full,” Dr. Joseph says. “You see the gyri. We’ll compare him to No. 2.”

POJ_1111_Image Perimeters【DFS】

/* Image Perimeters Time Limit: 1000MS        Memory Limit: 10000K Total Submissions: 9089        Ac...
  • huangshuai147
  • huangshuai147
  • 2017年04月17日 11:00
  • 232

POJ 1111 Image Perimeters 简单DFS Image PerimetersTime Limit: 1000MS Memory Limit: 10000KDescriptionTec...
  • yming0221
  • yming0221
  • 2011年01月18日 20:19
  • 1522

谷歌乳腺癌Detecting Cancer Metastases on Gigapixel Pathology Images

Detecting Cancer Metastases on Gigapixel Pathology Images 在高分辨率的病理图片上检测癌症转移 Liu Y, Gadepalli ...
  • Asun0204
  • Asun0204
  • 2018年01月27日 14:55
  • 126

ZOJ1047-Image Perimeters

Image Perimeters Time Limit: 2 Seconds      Memory Limit: 65536 KB Technicians in a pathology lab...
  • a664607530
  • a664607530
  • 2016年05月08日 15:50
  • 334

Teaching Machines to Read and Comprehend

关键词real natural language traning data, nerual model来源arXiv 2015.06.10问题针对阅读理解缺乏大规模训练数据集,构建了相应的数据集。同时...
  • u014300008
  • u014300008
  • 2016年10月09日 08:13
  • 904

手势估计- Hand Pose Estimation

1. 目前进展 1.1 生成模型      1.2 判别模型     1)J. Tompson, M. Stein, Y. LeCun, and K. Perlin. Real-Time Contin...
  • MyArrow
  • MyArrow
  • 2016年07月17日 19:12
  • 13033

POJ1111 Image Perimeters(dfs)

Image Perimeters Time Limit: 1000MS Memory Limit: 10000K Total Submissions: 8934 Accepted...
  • sUb_569
  • sUb_569
  • 2017年01月18日 16:54
  • 189


cursor:hand 与 cursor:pointer 的效果是一样的,都像光标指向链接一样,光标变成手行。cursor:hand :IE完全支持。但是在firefox是不支持的,没有效果。curs...
  • ran1004
  • ran1004
  • 2010年11月23日 11:28
  • 656

论文“Teaching Machines to Converse”阅读笔记

李纪为博士论文阅读笔记这是李纪为大神的博士毕业论文,趁热赶紧读一读,为自己科研道路指明方向。论文的下载地址是他的github1.1 背景知识介绍对话系统的分类及方法: Chit-Chat-orient...
  • liuchonge
  • liuchonge
  • 2017年11月28日 16:48
  • 470

Is teaching important?

        Is teaching important ?Well,of course it is.There was a time when the necessary knowledge co...
  • xteeq2009
  • xteeq2009
  • 2010年03月24日 10:21
  • 2767
您举报文章:In an Age of Images, Teaching Pathology by Hand