Digital Pathology Blog

Telemedicine to Receive Financial Boost

The Minneapolis Star-Tribune reports on the growth of telemedicine. The Mayo Clinic, for example, is getting into the act.

"The technology for telemedicine has been around for years, but there has been a catch: Insurers wouldn't pay for its use. They paid for visits to a doctor's office.

Now that's changing, so more patients are likely to start logging in for their doctor's appointments."

The payment hitch has prevented widespread use. But starting in January, the Centers for Medicare and Medicaid Services will pay for some telemedicine services, including from nursing homes and community mental-health centers. Private health plans tend to follow Medicare's lead.

One promising use of telemedicine: It can be cost-effective for checking in with patients who need frequent oversight: $12,000 for a year versus $15,000 for a single hospital stay.

In case you follow this market and missed this in the past few days - news from BioImagene.  This was discussed and rumored for some months and with previous news about Siemens executive joining BioImagene not too unexpected.  No dollar amount on the press releases I saw but fair to say even a conservative amount brings total investments since March of this year into a handful of digital pathology companies that have publicly disclosed their funding well over $100 million.   

I think the numbers are staggering considering what this market looked like just a few years ago.  Now the questions are not about image size, monitor resolution, speed...OK some of them are...but much excitement about workflow and incorporating this into the pathologist's office and enabling him/her with these technologies. 

With GE and Siemens in the market, this is encouraging for all pathologists.  I suspect news coming from Phillips is not far behind.

"Digital pathology company BioImagene Inc. said Thursday it received an investment from Siemens Venture Capital.

Cupertino-based BioImagene did not disclose the amount of the investment, but said it follows a recent closing of a $26 million fourth round.

Semens Venture Capital, which has an office in Palo Alto, is the corporate venture organization for Siemens AG, one of the largest global electronics and engineering companies.

BioImagene's funders also include San Francisco-based Burrill & Co., St. Louis, Mo.-based Ascension Health Ventures, National Healthcare Services, Palo Alto-based Artiman Ventures and ICCP Ventures.

Digital pathology is the process of converting entire glass microscope slides into high-resolution, whole-slide digital images that can be viewed, managed and analyzed with a computer instead of a microscope."

Is image sharing "social networking" that should be blocked?

An increasing number of pathologists share images for illustration and/or query "What is your diagnosis?" on cases whether for help or other thoughts or as unknowns once a case has been worked up and diagnosed.  These include sites such as Flickr, Photobucket and others. 

One in particular I check out from time to time is from a pathologist in Texas by the name of Ed Uthman who has posted thousands of pathology images on Flickr in addition to personal photos.  I use his examples for my own lectures and to teach residents as the images are generally of high quality and good examples.

Recently someone else on the listserv tried going to the site to view the images and apparently was blocked from doing so according to this post:

"And so, I just got on the phone with our local eHealth department at our local regional health authority. And the woman on the phone said that the instruction is to keep Flickr and other image sharing websites BLOCKED.

There is nothing else I can say. I am extremely angry having went the regular route of reasoning with IT department bureaucrats and having eventually had to speak to concrete walls.

She said Flickr is for "social networking" and has to remain blocked. I explained and explained but to no avail.

For the other colleagues who talked to IT departments and got Flickr unblocked, you probably have IT personnel with one drop more sense and logic than we have here."

Fortunately I do not have this issue at my institution but have seen this kind of thing from other IT shops including other popular sites such as those for web-based e-mail and other "social networking" sites. 

I think our hospital IT colleagues need to be sensitive about uniformly blocking "social networking" sites that often have other significant deliverables that health care professionals can derive benefit from by sharing their knowledge whether it comes in the form of image sharing, opinions, questions or discussion about the challenging issues we all face on a daily basis.  These Web 2.0 tools are enabling and highly informative and should be recognized as such as supported when IT departments have concerns that are contrary to that. 

Role in HER2 in gastric cancer

Gastric cancer is the second leading cause of cancer mortality in the world and its management, especially in advanced stages, has evolved relatively little. In particular, no targeted modality has so far been incorporated to its treatment armamentarium. HER2 overexpression is increasingly recognized as a frequent molecular abnormality, driven as in breast cancer by gene amplification. There is mounting evidence of the role of HER2 overexpression in patients with gastric cancer, and it has been solidly correlated to poor outcomes and a more aggressive disease. Additionally, preclinical data are showing significant antitumor efficacy of anti-HER2 therapies (particularly monoclonal antibodies directed towards the protein) in in vitro and in vivo models of gastric cancer. As a result, several clinical trials are exploring in different settings and with diverse designs the potential of anti-HER2 therapies in gastric cancer patients.

A recent review from the Annals of Oncology discusses the issues with HER2 in gastric cancer. 

This is another example of a potential targeted therapy in the age of personalized, predictive and pre-emptive medicine and yet another role for the likely importance of image analysis in such cases.

Download her2_in_gastric_cancer_a_new_prognostic_factor_and_a_novel_therapeutic_target_annals_of_oncology.pdf

High def in the lab

An article describing high-definition misroscopy imaging solutions for diagnostics and teaching recently appeared in Advanced Imaging Pro.  The article was written by Gary Pitre of Toshiba Imaging Systems Division.

Is this as good as it gets?

Geneticist contends human evolution may be grinding to a halt

The Press Association (10/7) reports that Professor Steve Jones, of the University College London, says that "human evolution is grinding to a halt," adding that "this is as good as it gets." In fact, he "believes the mechanisms of evolution are winding down in the human race." Jones explained that "there were three components to evolution -- natural selection, mutation, and random change." He added, "In ancient times half our children would have died by the age of twenty," but "our life expectancy is now so good that eliminating all accidents and infectious diseases would only raise it by a further two years." Thus, "natural selection no longer has death as a handy tool."

Furthermore, according to Jones, "human social change often changes our genetic future," the U.K.'s Times (10/7, Belluz) adds. While "chemicals and radioactive pollution could alter genetics, one of the most important mutation triggers is advanced age in men." Jones pointed out that "every time there is a cell division, there is a chance of a mistake, a mutation, an error," and "for a 29-year old father [the mean age of reproduction in the West], there are around 300 divisions between the sperm that made him and the one he passes on." But "a drop in the number of older fathers will thus have a major effect on the rate of mutation."

The "third factor – randomness -- is also an important ingredient in evolution," as "small populations which are isolated can change at random as genes are accidentally lost," the U.K.'s Daily Mail (10/7, Derbyshire) notes. Yet, "as the world's population becomes increasingly connected, the opportunity for random change is dwindling." Jones concludes, "So, if you are worried about what Utopia is going to be like, don't. At least in the developed world, and at least for the time being, you are living in it now."

Medting creates global image repository

Here is another story about Medting from a eHealth Europe and how they are creating a global image repository that I have been working with them on.  The number of relationships and people sharing content continues to grow.  The article describes their business plan in a little more detail:

A Spanish medical informatics consultant is enlisting researchers and practitioners in an attempt to create what might be the first truly global online repository of medical images and multimedia for physicians.

“The idea is to create a Web community,” says Miguel Cabrer, president of Medting.com. The interactive site, formerly known as MDPIXX, also includes platforms for clinical exchange and telemedicine.

The temptation is to call Medting a kind of YouTube for medicine, but Cabrer, a former e-health co-ordinator for the Balearic Islands Health Service, sees the site as more of a multimedia version of Snomed. Medting keyword tags are based on Snomed CT terminology, which can automatically translate keywords for indexing in multiple languages.

The site is free for doctors, who can create public or private groups for peer review, teaching, research and clinical decision support. Users can share documents and images, post comments and even vote on clinical interpretations.

A telemedicine component, partially funded by a European Commission grant, allows Medting to offer free second opinions to clinicians in developing countries, says Cabrer, a member of the Healthcare Information and Management Systems Society (HIMSS) EMEA governing council.

“People are using it for reference at the point of care or to discuss cases among colleagues and residents (junior doctors),” says HIMSS board chair-elect Dr Barry Chaiken, a health IT consultant in the US who is advising Medting. “You can get feedback from other people.”

Intel is sponsoring an awards program at Universidad Rey Juan Carlos in Madrid. Users have the opportunity to rate cases posted on Medting, and the top-rated will win prizes, Cabrer says.

Medting also offers enterprise-level services to organisations, for a fee, including a white-label version of the site. Information and multimedia can be embedded into external sites.

Among the enterprises that have signed up so far are the Catalonia government, which is using Medting for telestroke care, and USP Hospitales and the affiliated USP International Services in Spain, which have chosen Medting for a corporate image repository, Cabrer says.

The site currently is fully rendered in Spanish and English, and is starting to add content in other tongues. A highlighted case this week is in French, from Medting partner Geneva University Hospitals in Switzerland.

The company also has a relationship with the Mayo Clinic in the US and currently is looking for a dedicated US distribution channel—as well as a venture capital investor.

FDA Warns LabCorp about use of OvaSure

http://www.fda.gov/foi/warning_letters/s6947c.htm

I thought this was an interesting commentary:

"FDA aside, what can we tell someone about a screening test that is supposedly 95% sensitive and has a false positive rate of 0.6% among healthy individuals?

Let us assume that the annual figure of 21,000 cases underestimates the number of women with prevalent cancer by 50%, so, for argument's sake, let's say there's 40,000 women with ovarian cancer (rounded off).

If there are 50,000,000 women at risk for ovarian CA (1/3 of our approximately 150,000,000 women), then 0.6% of 50,000,000 is 300,000. So 300,000 women without cancer will test positive.

Let us assume for argument's sake, that all 40,000 women with ovarian cancer have the same likelihood of testing positive as those who were tested by Yale (a stretch, but probably not by much), so 95% of these, or 38,000 women with ovarian cancer will test positive.

So, in a best case scenario, at the start of screening, just over 10% - 38,000 out of 338,000 women who test positive will have ovarian cancer.

What do you do with this information? Perform ovarian ultrasound?

If the 20,000 "undetected prevalent" cases that I added above have a lower likelihood of testing positive, then positive and negative predictive values will be lower.

If women with "other conditions" (the initial study must have used completely "healthy" individuals as controls) not ovarian cancer, have alterations in these proteins, then the 0.6% figure could rise, further lowering positive predictive values. This says nothing about whether "early" detection will reduce risk of death.

This is NOT to say that this cannot become part of a useful algorithm, but the best way to prevent tests such as these from becoming useful is the commercialize them early with exclusive licenses and dispense with longitudinal studies, etc. Unfortunately, in our economic climate, where the overriding concern is the performance of a stock over the next 15 minutes, it is likely that we will see more of this.  Kudos to the FDA (and another blow to those who say all regulation is bad)."

RFID tags minimize specimen labeling errors

A colleague of mine in pathology and clinical colleagues in the gastroenterology divison at Mayo have reported their experience with RFID tags for GI biopsy specimens in a high volume practice.  The news was picked up on several sites including US News and World Report.  Comments from the clinical champions of the project and a review of their experience is below.  This will also be discussed at the upcoming APIII meeting in a few weeks.

"The Gastroenterology and Colorectal Surgery outpatient endoscopy unit at our facility yields over 30,000 specimen bottles that are sent for pathologic review every year," study author Dr. Dawn Francis, a gastroenterologist at Mayo Clinic, said in a clinic news release. "Over the past several years, Mayo Clinic identified some issues with mislabeling of tissue specimens in the units. Most labeling errors have been due to either the wrong patient label or no label being affixed to a specimen bottle. As a result, a quality improvement initiative was created to reduce the number of specimen-labeling errors."

Mayo Clinic researchers drew their conclusions in a review of specimen-labeling errors at their facility in three months before and after the implementation of RFID specimen labeling there. Prior to the new system, 765 errors occurred out of 8,231 specimen bottles sent to the pathology laboratory for evaluation. The error total plummeted, to 47 out of 8,539 bottles, after the RFID plan was instituted.

"It appears that this quality initiative, with emphasis on correct data creation and transcription point reduction, has the potential to significantly improve our clinical practice," said Dr. Schuyler Sanderson, the pathologist involved in the research.

Introducing Case of the Month!

I am trying something new here.  A new case each month will be presented with hope of presenting a Case of the Week if there is interest in doing so.  Readers can respond with comments or diagnoses for a particular case and share their opinions. 

This involves using some JavaScript and HTML code recently posted on The Daily Scan.  You can select the fields and magnification as you choose.

I would be happy to post any interesting cases if you can provide a brief clinical history, whole slide images and diagnosis.

October's case involves a case I reviewed in consultation.  The patient is a 54 year old man who presented with abdominal pain and was found to have a large retroperitoneal mass involving the pancreas and transverse colon. 

Sections from the tumor showed conspicuous areas of well-differentiated liposarcoma (slide 3).  In addition, there were focal whorls identified composed of tightly, concentrically arranged, ovoid to spindle cells showing plump vesicular nuclei, eosinophilic nucleoli and scant amphophilic cytoplasm (slide 2).  One of the sections also showed areas of bone formation (slide 1). 

The diagnosis and a reference are below the images.

Enjoy! And thanks in advance for your comments!

Slide 2:

Slide 3:

Diagnosis:

Dedifferentiated liposarcoma with neurallike whorling pattern and metaplastic bone formation. 

Reference:

Nascimento AG, Kurtin PJ, Guillou L, Fletcher CDM. Dedifferentiated liposarcoma. A report of nine cases with a peculiar neurallike whorling pattern associated with metaplastic bone formation. AJSP 1998; 22(8):945-55.