Thursday, July 22, 2010

Science, Art & Deception of the Mind

In his investigative report with the title "The mark of a masterpiece" published July 12, 2010, in The New Yorker, David Grann informs us about the novel use of methods adopted from forensic science to authenticate works of art. Deception seems intricately interwoven with art. Grann's fascinating story informs us about the scientific method and the power of deceit. It tells us about people who become vulnerable to deception and exploitation on their quest for affirmation of their believe because of pride, ego and money. Establishing the provenance of a masterpiece mainly relies on the subjective expertise to date. However, the desire to discover a new masterpiece by a fabled artist may obfuscate judgment and reason. Definitive objective prove is, therefore, highly sought after.

David Grann portrays the art expert Peter Paul Biro. Biro gained distinction in the world of art for using cutting-edge scientific methods to successfully identify fingerprints of famed masters like Turner, Pollock and Leonardo on hitherto-unknown paintings and drawings. He employs spectral microscopy and computer-supported image analysis to derive his conclusions.

We tend to believe that we are able to establish veracity beyond any doubt with hard data produced by objective machines that do not posses vested interests in the outcome of the examination. However, as David Grann's investigation elegantly reveals, the interpretation of the data may entirely lie in the hands of the investigator. He takes sole responsibility of deciding whether the evidence is sufficient to prove the identity of the fingerprint in question.

Because of our fallibility, more than one expert commonly shoulder such decisions. Peter Paul Biro states that he developed his own imaging equipment, which is one of a kind. No other expert could be asked to verify his findings because they lack the experience with the revolutionary methods he devised. The burden of prove rests with him alone.

This situation puts Biro in a unique position of extraordinary responsibility. The scientific method insists on the reproducibility of findings, preferably by independent examination. Additional verification with different methods may be necessary. Peter Paul Biro is intending to use DNA profiles from hair samples for this purpose. As the only expert in his method, he must know that his career and future depend on his truthfulness. Any scientist's credibility is intimately linked to her/his integrity. Deliberate misrepresentation of evidence is a serious offense, destroying the most highly-respected reputation.

  • Anyone versed in fluorescence or confocal microscopy could quickly learn how to image fingerprints. You may not even need a microscope. The detective in the video uses a magnifying glass to examine them. A CCD camera with a macrolense on a solid stand will suffice. A a high-quality light source and fiber optics are needed to produce sufficient illumination of the specimen as well as light filters for the wavelengths at which the images need to be captured. For storage and processing we need a computer with frame grabber and software.  In total, the equipment may not cost more than $100,000.-. The software for the described image treatment, that is background correction and subtraction, contrast enhancement and erosion to lay bare patterns, alignment and overlay of images as well as subsequent differentiation can be achieved with Wayne Rasband's ImageJ. The application is java-based freeware and can be downloaded from Wayne's National Institutes of Health site. The site also provides helpful information on equipment. For statistical analysis, we could quantify the number of picture elements, also known as pixels, in which the compared fingerprints overlap and use pixel-by-pixel paired t-tests to assess statistical significance as measures of identity. According to Grann, Biro charges $1,000.- per day for his qualitative analyses of fingerprints. His judgment must account for most of the cost (07/22/10).
  • Bryony Jones reports in her post with the title "Museum discovers 'new' painting" published online by CNN on Mar. 23, 2012, on an excellent example demonstrating how improved X-ray spectrography helped authenticate a Van Gogh painting previously attributed to an imitator. The painting depicts a still-life of summer flowers exhibited in the Kröller-Müller Museum in Otterlo, the Netherlands. It had been examined with less powerful X-ray methods before, and it was known that there was another composition underneath. Using the new imaging method, experts unambiguously uncovered under the still-life two wrestlers entangled in a match known to have been painted by the famous painter. In addition, the method confirmed that the pigments in the paints were consistent with those used by Van Gogh at the time. Modern technology correctly applied helps (03/24/2012).
  • On rare occasions, identification seems straight-forward. The proud new owner of this Renoir rediscovered comments: “One man's trash is another (wo)man's treasure!” (09/12/2012):
Related Posts

Tuesday, July 6, 2010

The Human Genome: A False Promise?

Genomic medicine, also known as personalized medicine, has become a universal buzz word, perhaps to attract more dollars to biomedical research. Only two months ago the new Director of the National Institutes of Health and former Director of the National Human Genome Research Institute Francis Collins announced the age of personalized medicine in the national media. In short, the sequencing of a person's genome will become less costly and available to everyone in the near future. Based on our genetic makeup, doctors will be able to treat prospective illnesses before they have a chance to afflict us. This country's biomedical research establishment invested billions of dollars in this idea.

Then, three weeks ago the new Director of the National Cancer Institute, scientific adviser to the president, and former Director of the National Institutes of Health Harold Varmus stated in the news that human genomic research is only useful to science at present. Nicholas Wade's article entitled "A Decade Later, Genetic Map Yields Few New Cures" and Andrew Pollack's article entitled "Awaiting the Genome Payoff" were published in The New York Times on Jun. 12 and 14, 2010, respectively, in the wake of what must seem like an about-face at the highest level of leadership in U.S. biomedical research. In both articles, the short-term prospects of personalized medicine seemed less certain.

Alas, understanding the human genome in health and disease is doubtlessly spawning important progress on numerous fronts in medicine. The discovery of modified genes in cancer cells and the proteins they encode is being used to develop novel chemotherapies. In particular, drugs for targeted therapies are currently being tested that may counteract the averse effects of modifications in genes KRAS and BRAF in a variety of cancers. In yet another advance, we may be able to adapt our diet to our personal genetic makeup with a better promise of controlling our weight, diminishing our risk for cardiovascular disease.

By contrast, uncovering the genetic bases of mental disorders like autism and schizophrenia has proved difficult. The findings to date suggest that such disorders are the result of highly complex interactions of a multitude of molecular signaling pathways involving many genes. Whether genomic research can benefit therapies to treat such mental disorders remains to be seen.


  • Recent studies uncovered a mutation of the gene ALK in tumor tissue of people with non-small cell lung cancer. ALK encodes the enzyme anaplastic lymphoma receptor tyrosine kinase. The fusion of this gene with the gene EML4 renders a tyrosine kinase persistently active that prevents programmed cell death from curtailing neoplastic growth.  Aaron Saenz describes a potential drug treatment using this important finding in his post entitled "Crizotinib Targets Gene To Stop Lung Cancer Tumors in 90% of Treated Patients" on Singularity Hub dated Jun. 9, 2010. No doubt, the development of drugs specifically targeting such cellular malfunction is highly desirable. However, less than four percent of lung cancer patients carry the EML4-ALK fusion, and the drugs tested to date only slow the progress of the disease. Personalized medicine definitely needs more research, time and money to come to fruition (08/05/10).
  • Peter J. Boyer wrote an informative essay with the title "The Covenant" about Francis Collins and his goals as the new Director of the National Institutes of Health. The essay was published online Sep. 6, 2010, in The New Yorker (09/10/10).
  • Because targeted therapies are expensive and may produce varying results, we may be confronted with difficult choices. This conversation between Paul Raeburn and Leonard Fleck entitled "Looking at the Ethics of Personalized Medicine" broadcast yesterday on NPR's Science Friday informs us about the dilemma inherent in this issue (09/25/10).
  • Julie Steenhuysen's report with the title "Gene tests inadvertently exposing cases of incest" posted yesterday on Reuters provides a striking example for personalized medicine's potential of uncovering information that raises ethical questions (02/11/11).
  • According to Sharon Begley's article on Reuters with the title "DNA pioneer James Watson takes aim at 'cancer establishments'" published online today, the co-discoverer of the molecular structure of DNA, Nobel-laureate, and eminent proponent for the human genome project J.D. Watson now believes that our greatest hope for fighting cancer is the use of antioxidants. That is, boosting antioxidants may help prevent cancer and blocking their actions in tumor cells may help with a cure. His voice lends credence to those who cautioned against excessive expectations in the promise of personalized genomic cancer medicine (01/09/2013).
Related Posts


Friday, July 2, 2010

The Value of Education, Economically recently ranked U.S. colleges based on return on investment (ROI) over high school graduates, that is how much more college graduates earn than high school graduates on average over a span of thirty years minus the cost for college attendance.

The figures are calculated with voluntary reports from graduates who did not seek post-graduate education, and they extrapolate from the past. If a school has recently changed its offerings profoundly, the numbers may be misleading. The ROIs are adjusted for length of study, drop-out rate and inflation. Costs for interest accrued on student loans are not included.'s list is ranked by thirty-year net ROI. The top ten exclusively consist of private institutions of higher learning. Prestigious engineering schools are in the lead. This is not surprising, because engineers seek post-graduate degrees less frequently than graduates with other majors and draw high salaries. Ivy League members constitute half of this group.

More intriguing, however, is the ranking by annual ROI (table below). This number represents the average return per year as percentage of the cost of college attendance. Only one private school can be found among the top ten of this list. The rest are public schools. In-state students garner the best return for their investment. Taking the high cost of student loans for private institutions into consideration, good public schools undoubtedly offer the best deal economically.

Colleges Worth Your Investment - Top Ten by Annual ROI

School Name School Type Average Cost for College & *Fed. Loan Debt in '09 [$] 30 Year ROI [2010 $] Rank

Annual ROI [%] Rank

Georgia Institute of Technology Public (In-State)   79,140.-
1,111,000.-   31 14.2   1
Brigham Young University Private   58,450.-
   797,000.-   77 14.1   2
University of Virginia Public (In-State)   74,410.-
1,038,000.-   38 14.1   3
College of William and Mary Public (In-State)   74,720.-
   895,000.-   51 13.6   4
Colorado School of Mines Public (In-State)   95,740.-
1,132,000.-   27 13.6   5
Virginia Polytechnic Institute and State University Public (In-State)   83,270.-
   854,300.-   62 13.1   6
University of Michigan Public (In-State)   84,690.-
   875,400.-   58 13.1   7
University of California at Los Angeles Public (In-State)   94,100.-
   961,200.-   47 13.1   8
University of California, Berkeley Public (In-State) 118,900.-
1,223,000.-   16 13.1   9
University of North Carolina at Chapel Hill Public (In-State)   61,830.-
   615,100.- 149 13.0 10
*On Aug. 25, 2010, I added the median federal debt loads of graduates entering repayment during the last fiscal year that the U.S. Department of Education recently released (see Jett Wells and Leah Finnegan's article for The Huffington Post with the title "The Best Colleges In America .. And The Amount Of Debt They Leave Students In (PHOTOS)" dated Aug. 20, 2010). The complete list in xls-format can be downloaded here. In comparison with the debt loads listed in the table above, graduates from some renowned private not-for-profit universities and colleges may owe less than $10,000.-, e.g. Williams College, Harvard University, and Princeton University, whereas graduates from others owe more than $30,000.-, e.g. University of Pennsylvania, Vanderbilt University, and Emory University. The debt may be low at some private not-for-profit institutions, because their students may qualify less for federal loans and receive more institutional financial aid. Conversely, greater debt means less aid and/or higher cost. According to Tamar Lewin's post in The New York Times with the title "Average College Debt Rose to $24,000 in 2009" published online Oct. 21, 2010, half the schools listed above leave their graduates with near average debt.

  • State prepaid college tuition and fees plans, also known as 529 college savings plans, are designed to pay towards tomorrow's college tuition and fees at today's price, plus provide a tax rebate. Considering the steep rise of tuitions and fees at public schools in the past decade alone, the plans seem a sensible choice for parents to keep their children's student loan debt low. However, as Jessica Toonkel and Jilian Mincer report in their post with the title "Prepaid college plans: shrinking options, rising risks" published online on Reuters today, some plans have lost on investments, are underfunded and may fall short in their promises, despite legislative guaranties. Regardless, the Virginia529 College Saving Plan appears to be one of the safest, providing an additional benefit to students at the University of Virginia, the College of William & Mary and other institutions of higher learning in this state (01/11/2012).
  • Alan Silverleib and Tom Cohen, report a new higher education initiative of the Obama administration aimed at improving outcome in their article with the title "Obama unveils plan to control college costs" published online on CNN today. The President announced the reform at the University of Michigan, which is on the top ten list above. Hopefully, this overhaul will benefit students at good public schools (01/27/2012).
  • Stephanie Simon reports in her post with the title "U.S. college ratings game set for shakeup" published online on Reuters yesterday that the recent scandal surrounding falsely reported SAT scores at Claremont McKenna College, a the prestigious private college on the West Coast, may deeply affect how institutions of higher learning are judged in this country (02/04/2012).
  • Jack Hough's insights in his article with the title "'Investing' in College? It Pays to Think Like an Investor" published online by The Wall Street Journal yesterday are well in tune with the thoughts in my post (05/05/2011).
  • Based on a student survey conducted by, Robert Franek, Laura Braswell and the staff of The Princeton Review compiled a list of the 300 best professors at US institutions of higher learning along with professor and school profiles. The list was published in softcover by Random House this year and is quite informative. For example, Harvard University (private; annual tuition: $34,976.-) contributes two professors to the list. By contrast, The College of William and Mary (public; in-state tuition: $13,132.-; out-of-state tuition: $35,409.-) is home to ten such professors (05/23/12).
  • Anna Prior and Matthew Heimer's article with the title "Which colleges help grads snare top salaries?" published online by yesterday completely affirms the basic conclusion of my post that good public schools deliver most bang for the buck (09/26/2012).
Related Posts


Thursday, July 1, 2010

Neurolaw & The Mind

NPR News broadcast three in-depth installments narrated by Barbara Bradley Hagerty in the past three days about the latest findings on the genetic and neural underpinnings of criminal behavior and how these insights may influence sentencing in court.

The product of the gene MAOA discussed in the first installment entitled "A Neuroscientist Uncovers A Dark Secret" is the enzyme monoamine oxidase A. This enzyme breaks down the neurotransmitters dopamine, noradrenaline and serotonin. Modifications of this gene have been associated with an elevated risk for criminality (Guo and others, 2008). Persistently increased levels of dopamine may result in volatile, aggressive behavior. The research discussed in the NPR broadcast was conducted with the principal investigator's own family members and awaits publication in a peer-reviewed scientific journal. The identification of a large number of genes associated with mental disorders like autism may serve as a reminder that it is unlikely that the defect of only one gene may cause a behavior that is the result of multiple complex nerve cell interactions in the brain.

The second installment entitled "Inside A Psychopath's Brain: The Sentencing Debate" reports on evidence obtained with functional magnetic resonance imaging (fMRI) that convicted psychopaths viewing scenes with content needing moral judgment show abnormal activation of cerebral cortex. The principal investigator recently co-authored a study of similar design in which read statements were used instead of visual scenes (Harenski and others, 2010).

The identification of genes with gene chip analysis as well as the identification of activated brain regions with fMRI that may play a role in criminality are probabilistic with attached uncertainties. The third NPR News installment entitled "Can Your Genes Make You Murder?" informs us about the impact of the science discussed in the prior installments on a recent court decision in the state of Tennessee as an example of the advent of neurolaw, that is the inclusion of neuroscience in criminal law.

As intriguing as the scientific observations on the workings of the criminal brain may be, the science remains in its infancy. Examining the brains of convicted felons does not permit us to determine whether the detected changes are the result or the cause of the criminal behavior. A genetic modification statistically associated with criminal behavior may point to a potential hazard. Yet, only the committed crime provides the necessary affirmation.

Therefore, judgment in court will remain based on the offenders' decisions and actions. The difficulty of our judgment is brought into focus, however, when we are confronted with people who are perfectly able to make informed decisions and know right from wrong in every-day life, but who commit the most horrible crimes when the conditions are right with a good chance of repeating such crimes in a similar situation.

Are psychopathic offenders capable of repentance? Will incarceration improve them? Is it possible to 'cure' their affliction or will they forever remain too dangerous to be allowed to live among us unsupervised again?

These questions are as old as mankind. Whether neuroscience will help us find better answers remains to be seen.

I expressed my ambivalent feeling towards neurolaw in a poem when the term evolved into a buzz word. The poem is written in magic ink. The writing takes some time to unfold:

Related Posts