Monday, April 2, 2018


Math like science is a kind of art. Parabola and sphere drawn with VTK.

Sunday, August 20, 2017

Building beautiful places

Think of a beautiful city or town or part of a city you’d go to on vacation? Paris, Barcelona, Kyoto, Japan, Boston’s Faneuil Hall area. Smaller towns like Rockport, MA.

What’s common about the places we’re thinking of?

They’re almost all old. And they’re lively places you walk around and experience up close.

Do you want to use an old cell phone? Old TV? old computer? old airplane? Why are we getting better at making just about everything except cities and towns?

Some people say the world hasn’t built a beautiful city in 100 years.

We even know how to build lively, beautiful places. Urban designers looked at and measured favorite cities and towns to figure out the rules for beautiful cities. They have local character, Barcelona, Kyoto and Boston’s Faneuil Hall area are distinctly different but all nice places. They’re compact enough to walk around and for using trains or buses. On the most lively streets the buildings are about two times taller than the width of the street. There are plazas up to 30 yards across so people can hail each other from the other side. Living, working, shopping, and eating are mixed together keeping the places lively throughout the day.

If we know how to build attractive, lively, towns why don’t we? We mostly build for cars now - not people. It seems functional but traffic clogged roads, parking lots, aren’t beautiful or lively.

Not building many new vacation like destinations makes them rare and expensive. They also tend to be tourist trap like toys and lack the schools, grocery stores, and offices for the real business of living. When you need to buy milk you drive to a grocery store in a strip mall.

Does building lively attractive places matter? Doesn’t living in a functional place and going on vacation to resort destinations work?

No. We’re not building enough new homes in places that have jobs, like Cambridge and Boston, because homeowners are blocking more development. Why?? because new development is often ugly and draws traffic - which often lowers nearby house prices. (The bane of homeowners.) We call this opposition to new development NIMBYism, Not in My Back-Yard and it’s skyrocketing home prices.

Boston metro median is over $425,000 and rising 8-10% a year. The places with good schools and close to work can be over $900,000. America has about 325 million people and Earth 7 billion, more than ever before, so we obviously need more homes by jobs.

We need to develop lively, walkable, healthy, transit centered places made for real living THAT people, even homeowners, will enjoy. This will help overcome the opposition to new development.

Building more homes and controlling prices in thriving places opens up opportunities to more people. People can come to work and pull themselves out of poverty. Young people could come and start careers. Investors and entrepreneurs would have money left over since it’s not sunk into houses, to build new businesses.

The Massachusetts state government has a transit oriented development program with incentives that encourage lively, walkable, train station centered development. We’re not seeing lots of this development because developers and homeowners are so used to car-centered development they don’t know anything else. Developers build for cars and homeowners resist all development.

New development done well can make our neighborhoods livelier, more useful, more enjoyable and enrich our lives -- in health and in wealth and open up opportunities for everyone.



Monday, December 19, 2016

Robust to the future of medical imaging

Machine learning can make medical image feature detection, annotation, segmentation, and classification much more robust to changes in the lab and clinic. Traditional deterministic image processing uses fixed pipelines of filters to denoise, simplify, and enhance. The deterministic filter pipelines are designed by experts in image processing who have spent time looking closely at the images, understanding the imaging system that took the images and understanding the goals of the imaging. Also any deterministic feature extraction process will always look at finite number of image features - even if that number is in the hundreds. What happens when looking at a new condition or disease that affects new never before examined features? What happens when the imaging process changed? When a dye is changed? When a MRI pulse sequence is changed? The deterministic pipeline fails in these cases and the pipeline itself has to be changed - meaning rewriting the program to some degree. Machine learning image processing systems only need retraining instead of rewriting making them more robust to future imaging and medical advancements and new areas of research.

Gene editing to see what happens when ...

CRISPR gene editing can increase understanding how diseases happen even before it can be used for treating diseases. CRISPR gene editing is new enough that testing and regulation will take so time before CRISPR gene editing is used in treatments at the doctors office. That won't stop CRISPR from making an impact on medicine soon. Gene editing is already building model organisms of disease - cell lines, lab animals such as mice, that mimic a disease helping researchers understand a disease. Gene editing also lets researchers change a gene to see what happens when something is changed. This transforms genetics and biology from observing genetic variation from the wild to experimenting with genes. Our understanding of the biology will increase rapidly and a deeper understanding of life and disease will let us develop new treatments whether the treatments are gene editing based or not.

Precision medicine even better than personalized

Precision medicine is even better than personalized medicine. We want treatments to work for as many people as possible. Personalized medicine tries to find who a treatment works for and what treatment works for a patient - a worthy goal. Precision medicine seeks to understand why and how a treatment worked or failed. From the why and how the treatment could be improved to work better from a broader, less personalized group of patients.

Data collections over individual studies

Collections of medical data from many studies are better for understanding disease than single studies. Conclusions linking genetic mutations to diseases in individual studies have often been wrong due to small sample sizes. By publicly aggregating the data from many papers together the data becomes larger and more powerful. Merging and comparing cross study data takes powerful normalization tools. Luckily our data normalization tools are getting more and more powerful. One tool is ComBat to remove batch effects in studies: available from Biocondcutor, for R,